EBS-AI Edition EN

Introduction

Integrating Artificial Intelligence (AI) technologies into a modern ERP information system is now a strategic asset for organizations seeking higher efficiency, flexibility and better decision-making.

AI not only works as an additional tool, but as a smart mechanism that leverages data to create real-time added value. At the same time, it transforms modern ERP information systems, turning them from simple data capture tools into smart business decision-making platforms. Through advanced data analysis, machine learning and automation algorithms, AI can identify patterns, predict future trends, and suggest optimal actions, significantly reducing human workload and error margins. This enables businesses to gain a substantial competitive advantage.

Processes such as stock management, financial forecasts, sales analysis and user support can be made more accurate, faster and tailored to the organization’s real needs. From stock optimization and improved financial forecasting, to user support and personalized reporting, the ERP evolves into an active partner that dynamically adapts to the needs of the organization.

From this release (AI Edition), the ENTERSOFTONE EBS ERP is entering a new era by integrating the use of AI into its core and main functions.

This document outlines the potential and benefits of integrating AI into EBS ERP, highlighting how leveraging intelligent technologies leads to better decisions, reduced operating costs and increased operational flexibility, laying the foundations for a sustainable and technologically advanced operational future.

The main objective is to maximize operational value and create a more intelligent, adaptive and future scalable information environment.

Basic concepts

Assistance

The ERP system supports the operation of multiple specialized AI Assistants, each of which is designed to serve specific operational needs and roles. Instead of a general, unified approach, the system architecture allows for the use of different assistants, who utilize a common technological base but are differentiated in their knowledge, context and purpose.

Having multiple assistants allows the ERP to offer targeted, more accurate and relevant responses, significantly improving user experience and enhancing the efficiency of operational flows. Each Assistant can be activated independently or in the context of an AI scenario, forming a flexible and scalable AI ecosystem.

The number of available AI assistants in this release is specific and clearly defined, covering key and critical operational scenarios. However, the system architecture is designed with scalability in mind, allowing it to be enriched with new, more specialized assistance in subsequent versions.

As the AI platform remains dynamic, it will be constantly enriched with new, more specialized Assistants, meeting the ever-evolving needs of organizations.

AI Scenarios

An AI scenario is a predefined, intelligent workflow within the ERP system that utilizes AI technologies to analyze data, make decisions, and propose or perform actions in an automated way. Essentially, it describes when, how and why AI is activated, as well as the operational result it aims to achieve.

Each AI scenario is linked to specific ERP business data and processes, such as sales, inventory, financial or customer support. Based on default criteria, historical data and rules, AI processes available information and generates predictions, alerts, suggestions or automatic actions, enhancing the effectiveness and consistency of decisions.

AI scenarios are designed to be scalable and adaptable to the needs of each organization. They can operate in real time or periodically, evolve through machine learning and integrate smoothly into the daily operation of users, turning the ERP from passive recording system into an active, intelligent business ally.

Image - AI Scenarios

Chat

The ability to talk with an LLM – using Assistants, of course – within the context of EBS ERP opens up new avenues, offering enhanced support, automation and quick access to information. At the same time, it makes your work easier and more efficient.

The AI Chat doesn’t function as a simple chatbot, but as a smart assistant who understands your ERP environment, your business data, and your current point of work within the system. At the same time, it transforms the way you interact with the ERP, allowing users to exploit business data through natural language.

The use of natural language significantly reduces the search time for various information and accelerates decision making. The system understands the work framework, provides direct answers, analyzes data in real time and suggests actions. This reduces the time of performing complex tasks, improves the accuracy of decisions and increases the overall productivity of the business.

In addition to providing information and analytics, the AI Chat also functions as an active business mechanism for performing tasks. Using natural language, the user can give orders and perform functions directly on the system, such as adding a new item, creating a customer, placing an order, scheduling an appointment or sending an email and more.

The AI is not just responding; it turns conversation into action, bridging the gap between information and execution. In this way, the ERP acquires a new, interactive mode of operation, where work is carried out directly through the context of the dialog.

In the EBS ERP environment, the user can utilize the AI in the following ways:

Scenarios and Standard users

By purchasing the AI Module, all users automatically join the Standard User category, which includes the use of ready-made AI scenarios.

C1 users have access to a set of AI scenarios designed to support daily tasks and improve productivity. At the same time, advanced scenarios are available, which offer more advanced automation, data analysis and decision-making support.

More specifically:

Functional module	Features
EBS AI Edition	Enabler at installation level for AI capabilities.
EBS AI Standard User	Basic AI capabilities. Recipes (no agentic capabilities). The Standard User mode is available per active user of the installation, and available Credits are for monthly consumption

	Standard User category users do not automatically have Agentic features and must be assigned an extra Standard Plan/medium Plan/Premium Plan package.

Chat & Agentic users

The ability to chat (with an Assistant) is only given to users with Agendic capabilities.

Specifically, the following categories are available:

Functional module	Features
EBS AI AgenticUser + Standard Plan	Increased AI capabilities ΑΙ recipes (global & context) & Tools with the Standard Agentic plan (monthly consumption per user)
EBS AI Agentic User + Medium Plan	Increased AI capabilities ΑΙ recipes (global & context) & Tools with the Medium Agentic plan (monthly consumption per user)
EBS AI Agentic User + Premium Plan	Increased AI capabilities ΑΙ recipes (global & context) & Tools with the Premium Agentic plan (monthly consumption per user)

According to the table above, the Chat feature is available in three classified levels (packages) of use, which vary based on the number of tokens available for consumption per month.

This scalable architecture allows for the flexible adaptation of the AI Chat to operational needs, ensuring a balance between cost, performance and operational efficiency.

The exact number of tokens available per package is determined and disclosed upon purchasing the Module and the relevant packages from the Sales Department of ENTERSOFTONE.

In summary, the Licensing model is illustrated in the following figure:

Figure 2 - Licensing model

Purchase of the AI Edition module is required.

Calling the assistant

This is the mechanism through which EBS communicates with the built-in AI Assistant from anywhere (scrollers, forms, olaps, dashboards, bits, automations), enabling real-time Artificial Intelligence capabilities.

Through this call, the system sends structured or unstructured data, operational context and instructions, taking as a result targeted and usable responses, suggestions or actions based on logic and understanding of AI.

Using “Open Assistant”, the ERP acquires a natural and direct channel of interaction with Artificial Intelligence. This allows complex data to be converted into understandable knowledge, user support with high-value suggestions, and smooth integration of AI into daily operational flows, without requiring expert technical knowledge from the end-user.

Figure 3 - Calling the assistant

Architectural model

The following illustration shows the architectural model of integrating Artificial Intelligence (AI) into EBS. The diagram illustrates the key building blocks of the solution, as well as the communication flow between EBS, the AI platform as well as the AI services and the external infrastructure involved in its operation. This architecture ensures the secure interconnection of systems and the exploitation of AI capabilities to support operational processes.

Steps to integrate AI into EBS ERP

Purchase the AI Module from the Sales Department of ES1 and purchase User Types.

In EBS > main toolbar > ? > About this application > License Activation, to integrate the Module to the system

Purchase the Connector for AI Studio from the EBS Cloud Store and create an Endpoint

Copy the Endpoint Header value in EBS in the Company parameter: Artificial Intelligence \ Cloud Store - Connector for AI Studio

In EBS > Tools & Settings > Artificial Intelligence > On-Boarding and AI user management:

• On Boarding, to connect the EBS to the AI Platform & Cloud Store.
• AI users management: (a) mass create AI users from the EBS users and (b) specify a Default Assistant for each user (optional)

Connect to the AI Portal with the invitation sent to the customer's email and assign packages to the AI users.

Upon completion of the above steps, the relevant Company Parameters in the Artificial Intelligence category concerning the integration with external services should be already filled-in.

More specifically:

• The parameters “Customer API key” and “Customer Subscription ID” refer to the connection of the installation to the AI platform
• The parameter “ES Cloud Store - Connector for AI Studio” is for the connection to the Cloud Store

• Finally, there is one more parameter in the same category, the "Core Utility Activation Serial Number" which corresponds to the activation serial number of basic AI functionality. It is used to post users and interact with the AI Portal.

Pay extra attention in the case of multiple EAS on the same database.

In particular, when multiple EAS (e.g. for load balancing purposes) are operating on the same database and for the same company, a separate Cloud Store key is required for each of them.

To specify all of the EAS on the same company parameter, the following format should be used:

{port1}:{cloudstore_key1}#{port2}:{cloudstore_key2}#...

Where:

• port: the RPC port, as defined in the ESClientConnect.xml file of the connection to the corresponding EAS
• cloudstore_key: the endpoint's Header Value, as calculated in the case of an EAS per base.

	The above method of registration applies only if there is more than one EAS on the same database.

A few additional actions will be described below, which the user can optionally perform in order to complete the AI configuration.

On-Boarding

The On Boarding process is the way to check that EBS is properly integrated with the two necessary platforms (the AI Platform and Cloud Store).

Connectivity and functionality are checked through this screen to ensure that the interface is successfully completed and that all required systems communicate correctly.

Figure 5- On Boarding

In detail:

• Integration with the ΑΙ platform: Check if the serial number is associated with the subscription. If the check-box is not enabled, it is possible that the purchase serial number of the AI Module differs from the EBS serial number. In this case, enter the purchase serial number in the relevant field and press "Test". If the serial number is identified, the check-box will be automatically checked.
• Integration with the AI Connector of the EBS Cloud Store: Includes filling in the corresponding Company Parameter used to connect EBS to the Cloud Store.
• If the company parameter is populated correctly, the check-box is enabled
• Otherwise, fill in the value in the nearby field and press "Ping".
• Microsoft Webview2: It concerns the installation Microsoft WebView2, which is necessary to correctly view the HTML pages that load inside the EBS.

If WebView2 is installed, the corresponding check-box is displayed as enabled. Otherwise, through the available link, you can go to the Microsoft page to install it.

Uploading AI content

Image 6 - Uploading content to the Assistant

Uploading AI content is one of the most important features of the system, as the Assistant you choose becomes aware of valid and organized information concerning your business.

Through this screen you can upload and manage content that is relevant to your company, such as internal documentation, links (URLs), company website sitemaps and other relevant information. This content is utilized by the AI to provide more accurate, targeted and operationally aligned responses.

After selecting the Assistant, select one of the following ways to import your files:

Import PDF	Import URL	Import File

Alternatively, you can select or drag & drop multiple files of different types for bulk import.

After selecting the files or URLs, they appear at the bottom of the screen in a list format. From there you can view or delete any item before proceeding.

In order for the Assistant to utilize and “learn” the content of the files you have imported, it is necessary to “train” it. After you have finished selecting the files, press the relevant "Train assistant" button to begin the training process.

Some examples of content that can be uploaded:

• Corporate rules and policies
• HR documents and procedures
• Building safety and management guides
• Internal operating manuals
• Product and service guides

By uploading this information, the Assistant becomes aware of valid and organized corporate content, acting as a “digital connoisseur” for the business. Therefore it can:

• Provide answers based on your official policies
• Reduce the need to search in scattered documents
• Enhance the consistency and correctness of the information
• Support faster training of new users

This feature elevates the AI from a general assistant to a specialized corporate consultant, fully aligned with your business structure, processes, and knowledge.

Managing AI Users

The User AI Management screen (EBS > Tools and Configuration > Artificial Intelligence) is used to manage the ERP users who will leverage Artificial Intelligence (AI) capabilities. Through this screen, the administrator can search for ERP users and enable them on the AI Portal, assign them AI capability packages and set the default AI Assistant they will use.

Figure 7-User AI Management Page

Summary statistics are displayed at the top of the screen:

• Registered on the AI Portal: The total number of users enabled to use the AI services.

• W/ assigned subscription package: The number of users assigned an AI capability package.

• With default AI Assistant: The number of users that have a default Assistant.

These data are automatically updated after each change.

Above the user list there are filters available, to easily identify users. These support:

• Searching for an ERP user by their user name or code.

• AI Assistant filter, to show only users who have a specific Assistant.

• AI Portal status, to display only registered or non-registered users.

• Only w/out Assistant, to easily identify the users that do not yet have a default Assistant.

The central area of the screen shows all ERP users.

The following information is presented for each user:

• User name

• ERP user code

• ERP role

• Default AI Assistant

• AI Portal status

• Date of creation of AI Portal account

One or more users can be selected using the check-boxes to perform mass actions.

Finally, the screen provides the following basic functions:

Adding a User to the AI Portal

The Add to AI Portal option enables the selected ERP users as AI users.

The process creates the corresponding account on the AI platform, connects the EBS user to the AI services and allows access to the application’s AI features.

Alternatively, the user can be created directly through the AI platform itself.

After the successful registration, the status of the user changes to Registered.

	An email address must be set for the EBS user.

Assigning a default AI package

With the Assign default package option, you can assign to one or more users the AI features package they will use.

The package specifies the available user attributes, such as: access to AI assistants, use of AI tools, available features depending on the subscription.

The feature supports mass assignment to the selected users.

Mass Assigning an Assistant

The Mass Assign Assistant option allows you to set the same default AI Assistant to multiple users at the same time. This feature is particularly useful during initial setup, or when the Assistant needs to be changed to a user group.

This way, whenever a user starts a new conversation, the Assistant that is optimally tailored to their role, department, and task type is automatically activated. This ensures targeted support, more accurate responses and faster service.

Integration with the ΑΙ Platform

Image 8 - Integration with the AI Platform

The AI integration feature allows direct access to the ES1 Artificial Intelligence Platform, through which critical AI operating elements can be managed and monitored.

Through the dedicated management panel, basic information and supervision tools are provided, offering complete transparency in the use of AI services.

In particular, you can monitor in real time:

• The total available credits
• The credits consumed
• The remaining available credits

This feature allows for:

• Effective cost control
• Planning and managing the use of AI resources
• Direct consumption overview per period
• Optimized use of the AI functions
• Purchase of additional user types and/or credits

In this way, the company maintains full control of the use of Artificial Intelligence, ensuring transparency, predictability and rational management of available resources.

At the same time, it is possible to purchase additional credits instantly, offering complete flexibility and seamless continuation of the AI capabilities.

As your business needs evolve, you can enhance available capacity at any time, without interruptions, delays or complex processes. The process is simple, fast and fully controlled by you.

With the option of buying credits:

• You can ensure that the AI Assistants remain continuously operational
• You can scale usage depending on the volume of work
• Periods of increased activity are supported
• You can maintain full cost and consumption control

Your business thus acquires a flexible and scalable AI utilization model, adapted to its real needs – today and in the future.

Updating the AI Assistants

The system’s AI assistants use your business data to answer questions and help you in your daily work. To keep them accurate and useful, they need to be periodically “updated” with the system’s newest knowledge structures and AI data models.

The feature “Update AI Assistants” renews the knowledge of ERP’s embedded AI assets based on the latest operational data and available system information structures and in its execution, analytical database views are created or renewed

This process is used so that the AI assistants can:

• provide more accurate answers,
• perform data analyses,
• use information from all ERP subsystems,
• improve the quality of business insights.

Image 9 - Integration with the AI Platform

	Running this process is recommended after upgrading the EBS or periodically, to refresh the AI structures.

AI visibility

Central AI call

The AI option is found in the main horizontal toolbar of EBS.

If no Assistant has been set for the Logged-in user (see Managing AI Users), all available Assistants are displayed and you can simply select one of them to open their interaction screen (Chat).

Otherwise, the CHAT screen is displayed directly.

Image 10-AI Assistants

The Assistants are specialized digital assistants. Each of them has a specific role (e.g. Sales, Finance, Marketing).

Each Assistant:

• holds specific knowledge
• has access to specific data
• can perform actions within the ERP

Keep in mind that each Assistant has specific knowledge in their training and engagement fields.

In the 1^st version, the following Assistants are available:

1. General Advisor (Sofia)
2. Commercial Director (Constantinos)
3. Marketing Specialist (Angeliki)
4. Financial Director (Thalis)
5. Supply Assistant (Hermes)
6. Documentation Assistant (Cyrus)
7. Company Assistant (Anna)

Sofia, the General Advisor, functions as a horizontal, general-purpose assistant, capable of responding to wider prompts, guiding users, and linking information from different parts of the ERP.

Constantinos, the Commercial Director, is trained on customer and supplier financial data and supports commercial processes such as sales, offers and customer relations, providing recommendations based on historical data and current trends. He knows their items and stocks in the company’s warehouses.

Angeliki, the Marketing Specialist, focuses on actions related to visibility strategy, campaign analysis and content creation, leveraging market and performance data to support targeted decisions.

Thalis, the Financial Director, helps monitor and analyze the company’s financial data. He provides information on revenue, expenses, profitability and cash flow, supporting strategic decision-making. He can identify economic trends, suggest improvements in cost management and provide a clear picture of the company’s financial situation in real time.

Hermes, the Supply Assistant, supports the management of the company’s logistics by helping to plan supplies, track stocks and coordinate orders. He provides information on product availability, identifies potential power supply problems, and suggests actions to optimize material flow and reduce delays and costs.

Also, Cyrus, the EBS Assistant, is specialized in understanding the structure and functionality of EBS, providing immediate support and explanations regarding processes, settings and technical capabilities. He is trained on the full documentation of the application, the release documentation, ESBooks and technical instructions for the use of the basic tools of EBS such as scrollers, automations, business rules, cubes, etc.

Finally, Anna, the Company Assistant knows the Corporate Documentation, and operates as a documentation assistant, by providing clear and concise information about the corporate documentation. She helps users quickly find policies, procedures, guidelines and internal documents, ensuring easy access to accurate and up-to-date information across the organization.

Finally, Watson, the IT Administrator knows is a digital IT Administrator that supports ERP management, monitoring and security and knows how to record all events that occurred in the database

Chat

On the AI Chat screen, each Assistant acts as a specialized digital assistant, trained on specific ERP data and business matters (e.g. Finance, Warehouse management, Sales, Supply management).

The Assistants support:

• Default questions (ready-made prompts)
• Free questions in natural language
• Actions such as:

• creating a new customer
• creating an order
• sending emails

By selecting an Assistant, dynamically suggested questions appear, which utilize its cognitive field and guide the user in meaningful and targeted analyses.

These suggestions:

• Are based on the data that the Assistant “knows”
• Focus on critical indicators and operational scenarios
• Enhance decision-making with quick access to insights
• Reduce the time of asking complex questions

Picture 11- Chat with Assistant Constantinos

The user can either select one of the suggested questions or formulate their own prompt in natural language. The system recognizes the context and enables appropriate data sources, providing immediate, documented and operationally usable responses.

In this way, the AI Chat is transformed from a simple search tool into an interactive mechanism of analysis and strategic decision support.

	Simply click on the AI option on the main toolbar to open the Chat on the right side of the screen. Alternatively, use CTRL + click to open the Chat in large size and on a separate page, in the center of the screen.

The main screen of the Chat comes with various built-in features to serve this purpose.

Figure 12-Explanation of chat icons

The title of the chat's tab is automatically updated based on the first message of the user, facilitating the identification and management of multiple conversations. In case of scenario use, the scenario title is displayed.

While the AI Assistant is processing a request and you are waiting for the response, useful tips and best practices are displayed, providing the user with additional information and making creative use of the waiting time.

13-In-chat tips

At the bottom of the Chat the “Consumption data” is available from the Info button. It displays the following information:

• Subscription package
• Available Credits
• Total credits
• Available balance (%)
• Next renewal

Finally, the Info button turns red when the available credits reach the preset limit, in order to give the user a low credit balance notification.

Prompts

Selecting an Assistant displays dynamically suggested prompts, related to their field of knowledge. The default prompts guide the user in targeted analyses and correspond to implemented tools of the system, which provide answers to critical operational prompts of daily or periodic use.

These questions are linked to default tools, which implement specific business logic through data retrieval mechanisms and transactions to and from the installation’s database.

The collection of prompts will be continuously enriched in the following publications, in order to meet the increasing needs of users and business processes.

In addition to the default prompts, the Assistant can receive and answer freely formulated prompts from the user through natural language, about a wide range of system information, such as business data, financial data, results and general information about the operation of the business.

Operational Areas of Assistant Knowledge

Indicatively, some of the knowledge of the Assistants are as follows.

Constantinos / Commercial Director

Constantinos is specialized in Sales, Purchases, Cash, ERP Stock

Sales

• turnover per year, month, items, customers and comparisons between these
• general turnover,
• sales by period,
• sales by customer, item or company,
• sales KPIs,
• trends and differences,
• sales analytics by color and size,
• sales analytics by lot,
• pending or open sales orders,
• customer offers.
• analyzes product data and calculates their turnover ratio.
• analyze Sales Growth
• analyzes financial data for the company’s salespersons, such as sales, turnover, profit and their general performance
• analyzes customer behavior.
• identifies whether there is a risk of a disruption of its cooperation or a decrease in its purchase volume (Churn Analysis) for certain customers. detects early signs of reduced engagement or loss of cooperation.
• runs RFM Analyses for a customer or more (based on their Recent Purchase, Purchase Frequency and Monetary Value). Evaluates the customers based on this.
• runs Days Sales Outstanding analyses (DSO) for one or more customers

Purchases

• market analysis,
• supply cost,
• supplier performance,
• pending purchase orders,
• expenses,
• supply statistics,
• supplier balances

Collections and Payments

• analyzes customer collections,
• supplier payments,
• bank movements,
• liquidity accounts,
• cash transactions

Customer and Supplier Balances

• customer balances,
• open documents,
• overdue claims,
• outstanding supplier invoices,
• balances per company or branch.

Warehouse and Stock

• stock balances,
• Product availability,
• stocks per branch,
• stocks by color and size,
• warehouse movements.
• Makes a stock replenishment proposal for a specific item, based on historical sales data, current inventory levels and supplier delivery times.
• Identifies items that have stopped moving lately, e.g. in the last three months and suggests ways to promote them to customers (new campaign)

General knowledge & actions

• recovery of trade account details,
• Trade account contacts or sites
• Item details,
• Item characteristics,
• relationships between items
• registers a sales order
• registers a purchase order
• registers complaints from customers
• registers meetings with customers
• registers leads
• registers sales opportunities
• registers a new item, customer, supplier, person
• makes transitions of sales order to Invoice-Delivery Note
• makes transitions of Delivery Notes to Invoice
• registers payments to a suppliers
• writes emails
• translates text

Examples of prompts:

• “What were the monthly sales?”
• “What items have the biggest growth?”
• “Which offers are still open?”
• “Which products are low in stock?”
• “Show the available products by color and size.”
• “Which suppliers have the highest market value?”
• “Which purchase orders are pending?”
• “What is the net cash flow of the month?”
• “Show me the largest debt movements.”
• “What proceeds have been made today?”
• “Which payments to suppliers are pending?”
• “Which customers have outstanding balances?”
• “Which suppliers invoices remain unpaid?”
• “Compare last year’s and this year’s turnover and tell me the top 10 customers”

Thalis / Financial Director

He is specialized in the financial and accounting analysis of ERP data

Financial Management and Accounting

• accounting records,
• debit or credit,
• account movements,

• cash and bank details;
• cash flow,
• financial indicators,
• accounting documents,
• general accounting entries,
• account charges and credits,
• entry calendars,
• accounting entries per period,
• aggregated financial data,
• changes in accounts,
• trends and comparisons of financial data.
• account balances of General Ledger Accounting,
• balances by company,
• balances per branch,
• short-term liability analysis,
• balance comparisons between periods,
• changes in financial figures.

myDATA support

• reconciles myDATA classifications.
• compares ERP classifications with myDATA classifications,
• detects VAT classification discrepancies,
• supports pre-populated VAT statements,
• analyze myDATA transmission data,
• identifies discrepancies between ERP and myDATA.

Sales

• turnover per year, month, items, customers and comparisons between these
• sales analytics by color and size,
• sales analytics by lot,
• pending or open sales orders,
• customer offers.
• general turnover,
• sales by period,
• sales by customer, item or company,
• sales KPIs,
• trends and differences,

Purchases

• market analysis,
• supply cost,
• supplier performance,
• pending purchase orders,
• expenses,
• supply statistics,
• suppliers balances.

Collections and Payments

• analyzes customer collections,
• supplier payments,
• bank movements,
• liquidity accounts,
• cash transactions

Customer and Supplier Balances

• customer balances,
• open documents,
• overdue claims,
• outstanding supplier invoices,
• balances per company or branch.

Warehouse and Stock

• stock balances,
• Product availability,
• stocks per branch,
• stocks by color and size,
• stocks per lot,
• warehouse movements.

General knowledge & actions

• recovery of trade account details,
• returns contacts or sites of trade accounts
• item details,
• Item characteristics,
• relationships between items
• registers a sales order
• registers a purchase order
• registers complaints from customers
• registers meetings with customers
• registers leads
• registers sales opportunities
• registers a new item, customer, supplier, person
• makes transitions of sales order to Invoice-Delivery Note
• makes transitions of Delivery Notes to Invoice
• registers payments to a suppliers
• writes emails
• translates text

Examples of prompts:

• “Show me the monthly general accounting entries.”
• “What were the biggest charges of last year?”
• “Account balance 54.00 per branch.”
• “Credit comparison between this year and last year.”
• “Which calendars have the most entries?”
• "General ledger monthly analysis.”
• ‘Balance of ledger accounts per branch.”
• “Short-term liabilities this month.”
• “Changes in account balance compared to last year.”

Hermes / Supply Assistant

Hermes is specialized in the management and analysis of production, inventory, production planning and supply chain data

Production Analysis

• production quantities, production costs
• production efficiency, production indicators (KPIs)
• trends and production comparisons
• production by period, production by product or category.
• creation of purchase orders for raw material shortages
• raw material consumptions,
• material consumption,
• consumption cost, consumption quantities,
• changes in consumption,
• open production orders,
• outstanding quantities to be produced,
• planned production orders,
• shortages.
• stocks raw materials, semi-finished product, finished products,

Work centers and Capacity Planning

• production work center analysis,
• availability per work center,
• work load per work center,
• available working hours,
• degree of utilization for production resources.

General knowledge & actions

• recovery of trade account details,
• returns contacts or sites of trade accounts
• item details,
• Item characteristics,
• relationships between items
• registers a sales order
• registers a purchase order
• registers complaints from customers
• registers meetings with customers
• registers leads
• registers sales opportunities
• registers a new item, customer, supplier, person
• writes emails
• translates text

Examples of prompts:

• “What was the monthly production?”
• “Production cost per product for 2025.”
• “Comparison of production with last year.”
• “What kind of products have the highest production?”
• “Consumption of raw materials per month.”
• “Cost of product consumption.”
• “Consumption comparison between periods.”
• “What productions are pending?”
• “Which planned orders haven’t been turned into productions?”
• "Material shortages for the upcoming productions.”
• “Find a product with specific specifications.”
• “Compare two products.”
• “Product availability in the warehouse.”
• “Other quantities to be produced.”
• “Balance of raw materials per branch.”
• “Product stock available.”
• “Which materials are low in stock?”
• “Quantitative balance of an item.”
• “Which work center has the biggest utilization?”
• “Work center availability for next week.”
• “Capacity analysis per work center.”

Angeliki / Marketing Specialist

Angeliki is specialized in customer relationship management (CRM), data marketing analysis, lead management and sales opportunities.

CRM Analysis

• fpr CRM tasks,
• meetings,
• emails,
• customer complaints,
• follow-up actions,
• customer interactions,
• CRM activities per user or group

Leads

• number of new leads,
• conversion analysis,
• lead trends,
• analysis of leads sources,
• lead quality assessment,
• Lead comparison by period or campaign.
• sales opportunities,
• sales pipeline,
• opportunity stages
• chances of completion,
• sales forecast,
• conversion rates,
• expected revenue,
• trends in sales opportunities.

General knowledge & actions

• recovery of trade account details,
• trade account contacts or branches
• item details,
• Item characteristics,
• relationships between items
• registers a sales order
• registers a purchase order
• registers complaints from customers
• registers meetings with customers
• registers leads
• registers sales opportunities
• writes emails
• translates text

Examples of prompts:

• “How many CRM projects were created this month?”
• “Which customer complaints are still open?”
• “What meetings are scheduled for today?”
• “How many new leads were created this month?”
• “Which campaign brought the most leads?”
• “Comparison of leads to last year.”
• “Which sales opportunities are currently in progress?”
• “What is the total sales pipeline?”
• “Which opportunities have a high probability of closing?”
• “Forecast sales for the upcoming quarter.”

Sofia / General Advisor

Answers general questions

Answers for:

• Customer, item, supplier details
• Customer, supplier contacts
• Item features
• Relationships between items
• Customer, supplier addresses
• registers a new item, customer, supplier, person
• registers complaints from customers
• registers meetings with customers
• registers leads
• registers sales opportunities
• writes emails
• translates text

Cyrus / EBS Assistant

Cyrus is specialized in providing information, instructions for use and support regarding the operation of EBS ERP.

The Assistant’s knowledge comes from:

• user manuals,
• manuals,
• operating guides,
• technical documentation,
• Processes and instructions for use of the ERP,
• help documents and knowledge base content.

Its purpose is to support users in the daily use of EBS ERP and provide direct answers about system functions, processes and capabilities.

Anna / Company Assistant

Anna is specialized in the search, understanding and presentation of information derived from the company’s documentation and internal documents.

The Assistant utilizes knowledge bases and document repositories that include:

• corporate documents,
• internal procedures,
• company policies,
• organizational documentation,
• operating standards,
• contracts,
• technical documents,
• presentations,
• reports,
• and other company documentation.

Its purpose is to provide direct and concise information about the company and its internal documentation through natural language.

This Assistant's material and knowledge sources are managed via the "Upload AI Content” option in the Tools and Configuration > Artificial Intelligence menu.

Watson / IT Administrator

Watson is a digital IT Administrator that supports ERP management, monitoring and security.

This Assistant knows how to record all events that occurred in the database, such as login, log out, entry deletion, etc.

Watson also contains a record of all modifications made to the fields in the database tables (field history log).

His main skills are:

• Detection and logging of events (logs).
• Monitor database activities and history log.
• Abnormal activity detection.
• Search of logs and action history.

Examples of prompts:

• "Show me the failed login attempts”.
• “Who deleted this entry?”
• ‘Show me the history log of item prices“

	You can search for Items, Customers, Suppliers, and Persons using both their main items and their UDF fields, even with the custom names you have set in EBS

Actions

The AI can also perform certain targeted ERP actions that are usually performed by the user. Through the chat screen the user can:

• Register a sales order
• Register a purchase order
• Add a new person
• Add a new customer
• Add a new supplier
• Add a stock item
• Register a customer complaint
• Register an appointment with a customer
• Register a lead
• Register a sales opportunity
• Send emails
• Register a production order

In this way, the use of AI facilitates the management of daily tasks and saves time for the user. All available actions will be gradually enriched in each new version of the system.

AI scenarios

In the Business Intelligence > AI menu there is the central screen for managing pre-made AI scenarios. The scenarios are shown by Category and for each one the title and a brief description explaining its function are presented.

As mentioned in the “Basic concepts” section, an AI scenario is a predefined, intelligent workflow within the ERP system, which utilizes AI technologies to analyze data, make decisions and propose or perform actions in an automated way.

Each AI scenario is linked to specific ERP operational data and processes and based on certain criteria, historical data and rules, available information is processed and analyzes, proposals, forecasts are produced, always aiming at optimal decision making.

Image 15- AI scenario screen

The structure of the page consists of:

• some basic categories such as Items, Customers, Suppliers, Sales, Purchases, Cash, Production
• the pre-made AI scenarios of each category with their title and detailed description.

This collection will be continuously enriched in future publications to meet the increasing needs of users and business processes.

Below you can find in detail and by Category the default AI scenarios and their corresponding functionality:

Items

Product evaluation

Analyzes sales data and identifies which products perform best for the business

Profit margin by Item category

Analyzes financial and transactional product data to calculate and assess the profit margin by category of items and to support strategic pricing and product portfolio decisions.

Stock replenishment proposal

Presentation of smart stock replenishment proposals with purchases and orders to suppliers, utilizing data such as inventory, sales, calendarization, and customer preferences

Shortage Forecasting & Stock Optimization

Forecasts stock-out risks, identifies stock excess, calculates default re-order figures and improves cash flow and storage costs. Reduction of reserved capital. It helps to prevent lost sales, better purchase planning, data-driven decisions rather than empirical decisions

ABC item analysis

Categorizes warehouse items based on their contribution to key indicators such as turnover, profit or sales volume. In this context, the items are classified into three categories: A, B and C, where category A items represent most of the total value, category B for items with medium contribution and category C for items with the smallest.

It is used to better manage stocks, control profitability and make targeted decisions about trade policy and item prioritization.

Identification of Items at Risk (CHURN)

Analyzes the sales and behavior of all items, identifying products that show a decrease in demand, a declining commercial path or a risk of creating slow-moving stock. The analysis provides timely notifications, highlights the key reasons that affect product performance and suggests targeted business actions to improve sales and optimal stock management.

Customers

Customer churn risk

Predicts the risk of business customers walking out (churn risk), explains the reasons that lead to it and suggests customer retention actions.

Forecasting of delayed customer payments

Helps to identify customers that pay late, assess the severity of the delay and suggest appropriate actions to reduce risk and improve cash flow.

Days Sales Outstanding (DSO) analysis

Analyzes the customers’ DSO (Days Sales Outsourcing) index, identifies deviations and suggests actions to improve collectability.

ABC customer analysis

Categorizes customers based on their turnover. Customers are classified into three categories: A, B and C, where class A customers represent the largest proportion of total value (e.g. turnover), class B for customers with medium contribution and class C for customers with the smallest.

Helps the company identify its most important customers by supporting targeted decision-making in areas such as commercial policy, service and sales development.

Analysis & Aging of Accounts Payable Optimization

It leverages the aging of accounts payable data and identifies high-risk customers, proposes priorities for collections, provides a clear picture of liquidity, ensures smart prioritization of collections and contributes to early identification of potential risk.

Analysis of Customer Behavior

Get a full picture of your customers’ behavior and act before valuable opportunities are lost. Analyzes markets, transaction frequency and total customer value, identifying early signs of churn or growth.

The system automatically categorizes customers (e.g. high value, in risk, inactive) and proposes targeted actions for each case. Identifies CRM activity (calls, complaints, emails, appointments, etc.)

In this way, you enhance customer retention, increase their life-cycle value, and turn data into meaningful business decisions.

RFM customer analysis

Analyzes customers using the RFM methodology (Recency, Frequency, Monetary) and categorizes them according to their value and behavior. Identifies the most important customers, those with churn risk and those with growth potential. The result provides salespersons and Account Managers with targeted retention, reactivation, cross-selling and up-selling suggestions for each customer category.

Sales

Salesperson performance analysis

Analyzes the performance of salespersons based on available sales data and salesperson targets

Sales forecast

Predicts future sales based on historical data

Production

Analysis of production planning data

Based on production planning data, deviations, congestion and optimization opportunities are highlighted and the optimal management of the production process is supported.

Production scheduling analysis

Production scheduling is evaluated, congestion is detected, delays and ineffective sequences identified and the optimization of production processes is supported.

Control of planned production orders

The AI monitors planned production orders, detects delays or deviations, and supports optimization of production flow.

Use of production work centers

The AI analyzes the operation of production work centers, detects loads, delays, and periods of high performance and supports the optimization of production processes.

Identification of Production Shortages

Identifies last-minute shortages related to orders pending production. It takes into account pending productions, pending consumptions and current stocks.

Overall Equipment Effectiveness (OEE)

The AI calculates and analyzes the OEE (Overall Equipment Effectiveness) for each work center, breaking it down into its three components: Availability (of equpiment), Performance (production speed) and Quality (of outflows). It identifies the main sources of loss — whether that is breakdowns, reduced speed or quality issues — and provides a comparative picture between centers and time periods, supporting maintenance, process improvement and equipment investment decisions.

Liquidity

Company Cash Flow Forecast

Based on recent financial data such as collections and payments, confirmed input and output forecasts as well as planned inflows and outflows, it makes a forecast for the future state of the company’s cash flows

Payment Optimization

It analyzes in real time the current cash balance, outstanding liabilities and scheduled inflows, creating an optimal payment plan for the selected time horizon. It identifies the optimal timing for payments of liabilities, so as not to generate a negative balance, taking advantage of the planned inflows.

Accounting

Fiscal Year Result Forecast

Based on accounting data, it observes revenue trend, expense trend and average arrival rate and calculates an estimated profit/loss of the year-end.

Identification of Financial risks, Balance Control & Accounting Anomalies

It checks if there are suspicious deviations in accounting records, detects sharp changes in accounts and abnormal balances, finds accounts with zero traffic but high balance, as well as possible accounting errors

Company

The Morning Bried

Morning Brief — your day in 60 seconds.

It makes a brief but comprehensive update on the current day. It synthesizes useful information and presents a narrative with prioritization.

Start every morning with a full picture of the business — without looking anywhere, without opening any reports. The AI is the trusted partner waiting for you every morning — he has already done the research and only tells you what counts.

Overview of the Company's Management

Presents in a single snapshot the sales and financial data of the company, combining sales and revenues, cost of purchase and production as well as the overview of cash. Offers direct supervision of the overall situation and supports fast and informed administrative decisions.

Summary Financial Overview of the Company

Provides a comprehensive and immediately understandable overview of the financial situation of the company.

Analysis of the Annual Corporate Balance Sheet

Summarizes the key figures of assets, liabilities and equity. Provides a clear picture of the financial position of the company and offers informed decisions at an administrative and strategic level.

P&L analysis

Analyzes the P&L statement, summarizing the company’s revenue, costs and profitability. Highlights trends and divergences, supporting informed decisions to improve economic performance.

Payment policy control

Evaluates the implementation of the company’s payment policy by monitoring payments to suppliers and partners. Highlights delays, divergences and optimization opportunities, supporting more efficient cash flow management and compliance with internal processes.

Company operating indicators

Analyzes basic operational data of the company, presenting statistical performance indicators in a concise and understandable form. The user acquires a direct overview of the operation of the company and can deepen the indicators and identify trends or deviations.

Application user behavior

Gain full control and visibility into your ERP usage with the intelligent user behavior analysis. Based on the History log and the Important event log such as login, log out in the application, entry deletions, identifies unusual actions, potential risks and errors in real time.

Through understanding usage patterns, it highlights discrepancies that may indicate problems or opportunities for improvement.

Thus, you enhance security, reduce risk and ensure the smooth and efficient operation of the business.

Complaint analysis

It aims to group and rank the complaints received by the company. Classifies them into the appropriate category, evaluates the severity of the issue, proposes a priority level and corresponding response time. At the same time, it evaluates the customer’s emotion, detecting levels of dissatisfaction, tension or urgency, identifies potential double or recurring complaints and links them to each other, creating a single picture of the problem. These complains feed a dynamic analysis layer that identifies recurring problems, emerging trends and points of improvement in the company’s processes.

Customer meeting analysis

Automatically analyzes meetings with customers and converts user notes into organized and usable information. It creates a brief summary, identifies the outcome and climate of the meeting, identifies customer needs and objections and suggests the next steps for the sales team.

At the same time, it evaluates the customer’s interest, identifies critical points and organizes the next actions of the team. The sales team gets a better view of each opportunity, reduces recording time and significantly improves follow-up and conversion process.

What amount of VAT do I owe?

It collects and processes data from sales and purchase documents and calculates the actual picture of the company’s VAT liability for the previous month.

In detail, it calculates:

-the total VAT of sales outflows

-the total VAT of purchase inflows

-the final VAT amount payable or the credit balance

-the VAT analysis per coefficient (24%, 13%, 6%, etc.)

-the VAT distribution by category of items or services

Item page

A series of predefined AI scenarios, designed to support commercial and financial decisions, are also available in the item form.

The system utilizes the available economic and historical data of the selected items (sales, calendarization, trends, profit margins, etc.) and produces dynamic analyzes and forecasts.

In this way, you can see the AI’s estimate of future sales, identify potential fluctuations in demand, and make more informed decisions about purchases, pricing policy, or inventory strategy.

Image - AI scenarios on the item page

Indicatively, some of the scenarios:

Best day for promotional actions

Historical sales data per product are analyzed to identify the days with a higher probability of high demand. Helps the sales department to time promotional actions and offers by item, maximizing sales and campaign efficiency.

Shortage forecast

A forecast is made as to when the item will be in shortage, that is, its stock will be exhausted. Takes into account demand and current balances of the item by crossing-checking them with certain safety levels.

Forecast of demand

Uses historical sales to predict future product demand.

Stock distribution

Presents the existing stocks per product and provides the distribution needs between branches or warehouse.

Helps optimize stock levels, reduces deficits and excesses, and ensures products are in the right place at the right time.

Detection of peculiarities

Detects any peculiarities in movements of the item such as sudden changes, large increases in the balance, entries of large sales or purchases (beyond what’s usual)

Item turnover

Based on recent historical sales and stock data of the item, it calculates how quickly it is sold and renewed within a specific period of time. It identifies whether the item is slow or over-stored, makes proposals for stock replenishment and optimization of stock levels. The aim is to reduce reserved capital, avoid shortages and increase overall efficiency of logistics operations.

Average inventory holding period

Analyzes the item movements from the warehouse tab, in order to calculate the average stock holding period. It assesses the efficiency of the stock and highlights cases of excessive stock. Based on historical data and recent trends, it provides practical suggestions for optimizing stock levels, improving traffic speed and making more efficient purchasing and sales decisions.

Stock replenishment proposal

Evaluates existing stocks and sales, proposing optimal quantities for replenishment. The aim is to maintain sufficient stocks, reduce shortages and optimize product availability.

Item valuation

Based on recent purchases and sales, the AI makes an estimate of the valuation price of the item with different valuation methods while comparing them.

Simulation of a sales price increase scenario

Simulates a scenario of increase in sales price/Forecast for sales, profit, cost

Customer page

In a similar way, a series of predefined AI scenarios are available on the customer page which support customer relationship assessment and management.

By selecting a scenario, the system analyzes the customer’s available data — such as purchase history, transaction frequency, financial data, balances, payment behavior and collaborative trends — and provides targeted assessments and suggestions.

The AI can highlight growth opportunities, signs of changing purchasing behavior, potential risk of customer churn, or suggestions for commercial actions, supporting more strategic and informed decisions.

Image 17-AI scenarios on the customer page

Indicatively, some of the scenarios:

Risk analysis

Assesses the reliability, payment behavior and commercial risk of the customer, with the aim of preventing risk and optimizing credit decisions. Identifies potential risks and provides evidence-based recommendations to Sales, Finance and Credit Control departments.

Purchasing power

Analyzes historical transaction data to draw conclusions about the purchasing power (purchasing power) of each customer. Market behavior (frequency, stability, price sensitivity) and product preferences are reflected in this analysis, aiming at usable information for sales, marketing and commercial policy.

Sales forecast

Predicts future customer sales based on historical data, trends, calendarization and customer behavior.

Churn risk

Predicts the risk customer churn, explains the reasons and suggests retention actions

Forecasting delayed payments

Predicts the risk of business customers walking out (churn risk), explains the reasons that lead to it and suggests customer retention actions.

Customer radar (anomaly detection)

Automatically detects unusual behaviors and deviations in customer movements, before they become a problem. It analyzes real-time data such as turnover, customer invoices by comparing them with customer history and identifies anomalies in the data — e.g. an invoice with unusual amount, a sharp drop in turnover. Each deviation is automatically categorized based on a risk level, so that the accounting department and management know exactly where to focus their attention. The result: fewer surprises, faster decisions and a stronger relationship of trust with the client list.

Analysis of Days Sales Outstanding (DSO) (Context in customer)

Analyzes the customer’s DSO (Days Sales Outstanding) index, assesses their payment behavior and identifies possible delays, arrears and deviations from the agreed payment terms. The DSO analysis shows the average number of days a customer needs to pay off their obligations, providing an insight into their credit behavior and risk level.

All the results of the above AI scenarios are displayed on the central chat screen of EBS. Information and knowledge don’t stop here, since the user can process them further, asking targeted questions for analysis and clarification. In this way, a great wealth of information is revealed, supporting decision-making and operational knowledge.

Bits: Items, Customers, Suppliers and Accounting

In all the Bits of Items, Customers, Suppliers and Accounting (Context & Global), the AI option “Analyze it” was added with the main objective of automatic and meaningful analysis of the results of each report. This function converts static data into usable information, making it easier for the user to understand and interpret it.

Image 18 - AI customer trial balance analysis

The system processes the data generated by the respective report, identifying trends, deviations, correlations and possible points of interest. Instead of being limited to simply displaying numbers and tables, the user gains direct access to conclusions and explanations based on the overall result of the report.

The AI analysis adapts to the content and type of each report, taking into account the business context and historical data. In this way, the AI functions as a digital analyst, supporting faster decision-making and enhancing the value of reports as management and control tools.

Olaps & Dashboards

All OLAP and Dashboards now support the AI “Analyze it” function, providing direct analysis of the displayed data through Artificial Intelligence.

Image 19 - Analyze it on the "Analysis of ABC customers” dashboard

EBS Olaps are a key multi-dimensional data analytics tool, allowing users to look at information from different perspectives, such as time, type, customer, supplier or account. Integrating AI analysis into cubes significantly enhances their value, turning complex information into readily understandable conclusions.

Through AI analysis, the system can automatically detect patterns, trends and deviations within the cube data, without requiring the user to manually examine multiple dimensions and combinations. The AI highlights important points, explains changes in data behavior, and helps the user quickly understand what matters.

In this way, ERP cubes cease to be merely a tool of analytical exploration and evolve into an intelligent analysis environment where Artificial Intelligence functions as an accelerator of knowledge and operational value.

Opening the assistant

In various parts of the application — such as entities, scrollers, Bits, and cubes — the “Open Assistant” option has been added.

Figure 20 - Calling the assistant

This option does not simply call one of the Assistants. In contrast, the Assistant receives all the full data and content displayed on the current EBS screen, so is fully aware of the context and works based on the actual operational context of the user.

The assistant call works:

• on the selected entries (selected rows)
• on the entries that appear after applying “Filter by selection” or/and “Filter by exclusion”

For example, on the customer card, by enabling “Open Assistant,” all of the data on the tab is fed to the AI, allowing the user to:

• ask targeted questions,
• discuss the data directly, and
• give immediate answers and advice and make real-time analyses.

Think of the Assistant as a smart partner who sees exactly what you see and guides you directly!

At the same time, the Assistant can identify trends, discrepancies or opportunities that are not immediately apparent, supporting informed decision-making. The analysis does not remain theoretical; it can be turned directly into actions within the ERP, accelerating daily processes and reducing the time from information to practice.

Calling an assistant is not just an additional feature of the system. It is the ERP’s transition to a new operating model, where data is not static but interactive, and information is instantly transformed into knowledge and action. AI is organically integrated into everyday work, transforming the ERP from a recording tool to an active, intelligent decision-making partner.

Automations

The AI capability “Analyze it” is also available in the automations, allowing the immediate use of Artificial Intelligence to interpret their design.

This function clearly presents the feasibility, functionality and operational value of each tool, while providing detailed technical information that supports its understanding and proper utilization.

Image 21- AI automation analysis

Scrollers

In the scroller design, the AI option "Analyze it" was added.

This function clearly presents the structure and information value of each scroller (reference), while providing detailed information that facilitates its understanding and proper utilization.

In particular, it includes listing the parameters and columns displayed, internal filters applied, expression calculations, critical points of interest, as well as a summary of the scroller.

Image 22- AI scroller analysis

Olaps

In the OLAP design, the AI option "Analyze it" was added.

This function offers a complete and understandable presentation of the structure and information value of each OLAP providing detailed information that greatly facilitates its understanding, control and proper utilization.

In particular, the parameters and columns that participate in the OLAP, horizontal and vertical analysis axes, the facts (output fields), the calculated fields, the internal filters applied, as well as the calculations and expressions used are displayed. At the same time, the AI can detect patterns and anomalies in the OLAP design.

.

Image 23 - AI Olap Analysis

Cognitive tools

Cognitive tools are software tools that can “think,” analyze, and process information in a way that mimics certain human mental processes. Unlike simple tools that run predefined commands, cognitive tools understand insights based on the data available to them.

The basic functions are:

• Spell-checking

• Text review

• Text reading

• Translation

• Optical character recognition (OCR)

• Analysis of emotion

EBS ERP has has always had integrated cognitive functions in all text fields. In its new architecture, these capabilities leverage AI by acquiring significantly expanded capabilities, greater accuracy and increased intelligence, enhancing overall user experience and efficiency.

At the same time, cognitive tools are integrated in the conversation with the AI assistants to significantly enhance daily work. They ensure consistent and professional formulation in line with corporate tone, automate error correction, provide instant translation and support smart features allowing users to save time and focus on higher value strategic tasks.

Presentation of the results

Based on the results obtained from the responses, the Assistants can create and present the content in different formats, such as text, tables, graphs, images, maps, presentations, texts or spreadsheets, depending on the user's request.

The display format is dynamically adjusted according to the type of request and content, offering more understandable and effective visualization of the information, facilitating the understanding of results and drawing conclusions.

Graphs

The AI Assistant is not limited to providing answers in text format. When the data allows, it can automatically create graphs and visual representations.

Through interactive diagrams, the user can monitor trends, compare measurements, analyze key performance indicators (KPIs) and easily detect significant changes in business data.

Presentations (pptx)

The AI embedded in EBS ERP can automatically create professional presentations at any time and upon request within a discussion. These presentations illustrate in an organized and understandable way the results of the discussion, present the key conclusions that emerge from the data and propose further actions and business directions, effectively supporting the decision-making process

For example, if during a chat the user asks for an analysis of turnover per District for the last year and this year, the AI can automatically create a complete professional presentation (PPT). The presentation includes comparative data, graphs, statistical analyses, key conclusions on the course of sales, as well as suggested business actions and strategies for the exploitation of results

Figure 28 - Displaying results in a Presentation

Excel

The AI can automatically create Excel files. These files organize and present the data generated by the discussion in a structured and easy-to-use way while facilitating the further processing, monitoring and utilization of the information by the users.

Figure 29 - Displaying results in Excel

PDF

The results of a discussion can also be recorded in PDF documents. These documents collect and present in a professional and structured way the information, conclusions and results obtained from the discussion or analysis of the data. At the same time, they can include tables, graphs, reports and action proposals, facilitating documentation, information distribution and effective support for the decision-making process.

Image 30 - Displaying results in PDF

Maps

It also has a built-in map and geographical display function, offering the ability to display addresses, points of interest and operational data on an interactive map. Through the use of geo-location technologies, users can monitor locations, organize routes and manage the daily operations and deliveries of the business more effectively.

For example, the user can ask the system to organize all the orders of the day and the AI to automatically create the most efficient order of visits to the map, optimizing the time and productivity of deliveries.

Other possibilities

The Chat provides the ability to record the chat's auxiliary file (log). This file can make a significant contribution to identifying and investigating possible errors.

When sending customer feedback about the use of AI to EBS, it is recommended to attach the relevant file to facilitate the analysis of the problem

The file is saved in folder: ESNoSync\AI_TroubleShoot_ChatTool

Figure 32 - Chat log

Security and Access Control

The access and data protection capabilities for each AI user are centrally defined through the AI Platform and EBS Privileges.

1. Access Rights Management on the AI Platform

Connect to the AI Platform through the option “AI integration”
(Tools and Configuration → Artificial Intelligence).

Then go to Package Management → Package Access Policies and create the Access Policies required for your organization.

For each access policy you can specify:

• The available scenarios (Recipes)
• The available Tools (Skills) per Assistant
• The available Knowledge (Data Agents) per Assistant
• The available Data Sources per Assistant

After the configuration is complete, assign the appropriate Access Policy to the AI users.

2. Protection of Sensitive Data through EBS Privileges

In addition, it is possible to restrict or hide sensitive information, such as:

• Turnover data, costs of goods sold and financial data
• Personal data of trade accounts that fall under the GDPR

The protection of this data is implemented through the existing EBS rights and roles mechanism, ensuring that each user has access only to the information permitted by his rights.

Let’s assume you want to hide the turnover.

A. Create a User Group

B. In Set privileges, select the user group and in the node: All protected options > Fields > (AI) Sales and choose Turnover and at the “Show” privilege choose: NO.

Let's say you want to hide your TRN and customers’ Phone numbers.

A. Create a User Group

B. In menu Tools and Configuration > GDPR procedures > Change field behavior, set the fields you want to restrict for GDPR reasons.

In detail:

• Select ESAICustomer for customer details
• Select ESAISupplier for supplier details
• Select ESAIPerson for person details

and configure the behavior of the fields you want.

C. Finally, when setting privileges, select the user group and in the node: Fields > (AI) Customer choose the TRN and the phone numbers and in the “Show” privilege select: NO.

Figure 34 - Setting privileges

3. Entry protection

The AI enables the creation of specific entries to EBS, concerning tasks and documents. This feature is determined by the rights granted to users through the EBS user groups. These rights are also fully respected by the AI, which applies the same access and action restrictions.

AI-powered workshops

The system supports the execution of automated workflows that utilize Artificial Intelligence to perform operational processes without user intervention. These flows monitor predefined operational events, analyze available data, make decisions based on the rules set and automatically perform the corresponding actions within EBS.

The aim of the AI Workflows is to automate repetitive processes, reduce registration time, limit manual actions, and improve overall user productivity.

Workflow: Automatic edit of incoming emails

This workflow tracks incoming emails, analyzes their content using Artificial Intelligence and identifies requests that require operational action. Depending on the content of the message, it automatically creates the corresponding entry without requiring manual creation by the user. This speeds up the management of incoming requests, reduces creation errors and ensures that each request is created and launched in a timely manner.

Initially, the AI analyzes each incoming message and categorizes it according to its content (specifically classifies it into one of the 3 categories: Complaint, Meeting Request, New Order). A detailed analysis of the e-mail content is then carried out to identify the operational action required.

When the AI recognizes with sufficient certainty that the message concerns:

• a request for a meeting
• a customer complaint
• a sales order

it automatically creates the corresponding entry in EBS by filling in the necessary items extracted from the e-mail content.

The operation is implemented through a Business Rule, which can be enabled or disabled by each installation, allowing each organization to decide whether it wants to fully automate this process according to its business needs.

	The Business rule is the AI Job in the ESTMInteraction entity and by default it’s Inactive. As email the system uses the task type (Task) with an international ES.EML code. This code can be modified per installation to adapt to the requirements and configuration of each business

10 good practices for AI Assistants

To get the most accurate and useful results from your AI Assistant during the chat, it is recommended that you follow the following practices when using it.

1. Be as specific as possible

The clearer your request, the more accurate the answer will be.

Less effective:

• Show the VIP turnover.
• Show me the sales.
• How much did we sell?

Preferably:

• Show the turnover of customers belonging to the VIP category for 2025.
• Show the sales per month for this year.
• What is the total turnover of customers in Attica for the first quarter of 2025?

2. Always indicate the time period

A lot of business data depends on the time period you are interested in.

Unclear:

• Show me the quotations.
• How many sales do we have?

Clear:

• Show me the sales quotations for the last month.
• How many sales were made in March 2025?

3. Use full business terms

Avoid abbreviations or internal terms that may have multiple interpretations.

Less clear:

• Show the VIPs.
• Show category A items.

Preferably:

• Show the clients that belong to the Major category.
• Show the items that belong to the category "Spare parts".

4. Start a new chat when the topic changes

The AI Assistant takes into account the history of the current chat.

When you completely change the subject, it is recommended that you start a new chat or clear the history.

Example:

• Chat about customer sales.
• Chat about warehouse stocks.
• Chat about financial data.

Using separate chats helps the Assistant better understand the new context and avoid incorrect associations.

At the same time, when a chat grows a lot, the AI needs to reprocess more of the history to maintain the context. This results in increased use of credits which can affect the speed and quality of responses. Create a new chat for new topics or separate tasks.

5. Avoid very large and complex chats

It is preferred to divide complex requests into smaller steps.

Instead of:

Give me the turnover per customer, then compare it to last year's, find the deviations, explain them and make me Excel.

Preferred:

• Give me the turnover per customer for 2025.
• Compare it to 2024.
• Find the customers with the most changes.
• Create an Excel file with the results.

Maintaining very large chats increases consumption of credits, as the AI takes into account all previous history before answering. For more efficient use, keep each chat focused on a specific topic and open a new one when a different task starts.

6. Clearly mention what you’re searching for

Specify whether you are referring to customers, items, suppliers, quotations, documents or other business items.

Unclear:

• Show me the top ones.

Clear:

• Show the 10 customers with the highest turnover.
• Show the 10 items with the most sales.

7. Ask for the desired result

Mention if you want a table, a chart, an analysis, or a file export.

Example:

• Show the results in a table.
• Create a chart of sales per month.
• Export the data to Excel.
• Create a report in Word.

8. Check the filters you are using

The more filters you give, the more targeted the result will be.

Example:

Show the sales of Major category customers in the Attica region for the first half of 2025.

9. Use natural language

No special syntax or SQL knowledge is required.

Examples:

• Who are my 20 best customers this year?
• Which items have the highest sales growth since last year?
• Which suppliers made no orders in the last 12 months?

10. Save frequently asked questions to Favorites

Frequently used prompts can be stored in the Favorites so that they can be performed instantly, without having to be typed again.

This is particularly useful for:

• Monthly sales reports.
• Comparisons of periods.
• Monitoring key indicators (KPIs).
• Frequent exports to Excel or Word.

Finally, we recommend that you use the evaluation icons

Evaluating AI responses through the available icons (positive evaluation) and (negative evaluation) is important for the continuous improvement of the system. Feedback helps identify the responses that are useful and those that need improvement, so that the user experience gets better over time.

AI analysis limitations

As EBS ERP enters a new world of learning and functional automation, it must be made clear that in any case, AI analysis is based on generative AI models and the data available, which means that the results may diverge from the actual business overview. For this reason, the conclusions and suggestions of AI should be evaluated and confirmed by the user before making critical decisions.

AI analysis provides support and clues, but its results may not always accurately reflect reality. For this reason, it is recommended that the results be monitored and used as an adjunct to decision-making.

Artificial Intelligence serves as a tool for supporting analysis, rather than a substitute for human judgment.

Results and forecasts may be affected by the quality, completeness or interpretation of the data, and therefore user verification is required prior to any business or strategic use.

PDF Version