In our recent AI Agents in Business Central webinar, we demonstrated the next frontier of ERP efficiency: AI Agents in Business Central. While many are familiar with AI assistants like Copilot, Microsoft is now moving toward an ecosystem of autonomous agents that can understand business context, make decisions within defined boundaries, and execute complex workflows with minimal human intervention.
For years, ERP users have spent significant time on repetitive data entry. With AI agents, the focus shifts from processing documents to customer service, business growth, and strategic decision-making.
How AI Works in Business Central: From Copilot to Autonomous Agents
AI in Business Central is categorized into three levels of autonomy:
- AI as an Assistant: Traditional Copilot features that summarize data and assist in decision-making.
- Human and Agent Teams: Where the agent acts as a “colleague” performing specific tasks, such as the Sales Order Agent and Payable Agent.
- Human-Led Agent Operators: A higher level of orchestration where humans define the boundaries and goals, and agents operate autonomously within them — ideal for complex, multi-step business processes.
How the Sales Order Agent in Business Central Handles Inquiries and Quotes
The Sales Order Agent is designed to solve the pain point of manual inquiry handling. It can:
- Understand Natural Language: Customers can send inquiries via email using informal descriptions rather than exact item numbers.
- Automate Quoting: The agent reads the email, checks inventory, and automatically generates a sales quotation.
- Convert Orders: Once the customer confirms the quote via email, the agent automatically creates the sales order in Business Central.
How the Payables Agent in Business Central Automates Invoice Processing
The Payables Agent automates the tedious task of manual invoice entry. Key features include:
- Zero-Training OCR: Unlike older systems, you don’t need to train the agent on where to find invoice numbers or dates, it understands document structure automatically.
- Vendor Recognition: It extracts data from PDF or image attachments and can even prompt the creation of a new vendor if they aren’t already in the system.
- Historical Learning: The agent improves accuracy over time by learning from previous transactions and item descriptions.
Why Human Oversight Still Matters in AI-Automated ERP
While these agents offer significant time savings with some scenarios showing 1.5 hours saved on email processing and 30 minutes on quotations, humans remain essential. The Human-in-the-Loop approach ensures that documents are reviewed for accuracy before final posting, maintaining financial integrity and giving teams full confidence in their data.
What Attendees Asked in the AI Agents in Business Central Webinar
1. To what extent can we customize these AI agents?
Currently, the two built-in agents (Sales Order and Payable) work out of the box and do not offer deep customization. However, Microsoft provides a “Build Your Own Agent” option, allowing organizations to create custom agents tailored to their specific business needs using Business Central data.
2. How does the Payable Agent handle matching Goods Received Notes (GRN) with invoices?
The agent supports three-way matching. If the vendor includes a PO number on their invoice, the agent traces the purchase order, compares it with received quantities, and allows for verification before the invoice is posted.
3. Is the Payable Agent available in all regions?
The Sales Order Agent is available globally. The Payable Agent is being introduced gradually across different regions and is expected to be widely available within the coming months.
4. Can the agent automatically post invoices or convert documents?
Currently, built-in agents are designed to create documents such as a draft invoice or quote. To automate final posting, we recommend using standard Business Central workflows or job queues.
5. How does the system ensure the accuracy of data processing?
Accuracy improves through a learning curve. While the first transaction from a new vendor may require closer scrutiny, the system improves as it processes more invoices from the same source. The built-in review step also ensures a human always verifies the document before it moves to the next stage.
6. Can these capabilities be extended to scanned delivery dockets or other documents?
While the standard agents follow specific workflows, additional automation such as reading delivery dockets or timesheets can be achieved using Azure AI Document Intelligence integrated with Business Central.
Getting Started with AI Agents in D365 Business Central
AI agents in Business Central aren’t just a feature update, they represent a fundamental shift in how ERP works. The move from manual processing to intelligent, context-aware automation is already underway, and the organizations that adopt it early will have a clear operational advantage.
If you’re ready to explore what AI agents can do for your business, Cetas is here to help. Whether you’re migrating to Business Central or looking to optimize your current setup with AI, our team is ready to guide you every step of the way.
Get in Touch With Us
Fill out the form below and we’ll get back to you as soon as possible!
