📅 Date: Feb 19, 2026
​​🕔 Time: 17:30 - 21:00
📍 Location: Raamstraat 7, 1016 XL, Amsterdam

Join us for an evening of practical case studies and networking. Assil Jwair & Bob de Witte will show how IKEA built a reliable conversational analytics tool with multi-agent AI, allowing business users to query complex data and get consistent, hallucination-free answers. Then, Fried Schölvinck, Agentic AI specialist, will demonstrate how to build agents with natural language in Open WebUI, enabling powerful features like agentic RAG and web search.

Additionally, you'll have the opportunity to connect with like-minded professionals over drinks and snacks before and after the presentations.

17:30 - 18:15: Welcome with food and drinks!
18:15 - 18:20: Opening notes
18:20 - 18:55: Talk 1 - Ask Your Data Anything: How IKEA Built Reliable Conversational Analytics with Multi-Agent AI by Assil Jwair, Lead Data Engineer at IKEA and Bob de Witte, Analytics Engineer at Xomnia
18:55 - 19:10: Break
19:10 - 19:35: Talk 2 - Building an open source agent creator in Open WebUI by Fried Schölvinck GenAI Lead at Xomnia
19:35 - 19:40: Closing notes
19:40 - 21:00: Networking/drinks

Talk 1 - Ask Your Data Anything: How IKEA Built Reliable Conversational Analytics with Multi-Agent AI
Assil Jwair and Bob de Witte will talk about how IKEA’s order allocation analytics team is leveraging LLMs to enable business users to query vast, complex data models through natural conversation. Beyond simple chat, we will discuss the architectural framework required to eliminate hallucinations, ensure consistency, and embed the expertise of data analysts directly into the tool. Discover how this approach accelerates decision-making and provides a scalable, reliable way to unlock insights.

Talk 2 - Building an open source agent creator in Open WebUI
In this talk, Fried Schölvinck will learn you how to build agents with natural language in Open WebUI, a popular ChatGPT-alternative. By leveraging the exposure of internal API's to the models via tool calling, modern LLMs are able to create new agents and even edit their own configuration, directly in the UI. Peak inside the source code of a mature AI framework and discover how to use that built-in python code to enable agentic RAG, web search and more. Fried is an agentic AI specialist and machine learning engineer with a background in Neuroscience and Artificial Intelligence. Motivated by the vast potential of AI and a pragmatic approach, he has pursued a career blending Machine Learning Engineering with consultancy.

Important Note:
⚡ Spots are limited! Please cancel your registration if you cannot attend.