How do we make AI accountable not only for what it does, but also for what it knows?
Good to know
Highlights
- 3 hours
- In person
Location
Askend
80f Tartu maantee
2nd floor 10112 Tallinn
How do you want to get there?
Agenda
The EU AI Act: Compliance Is an Engineering Problem
The EU AI Act is now rolling out, and it affects far more developers than most realize — anyone building with LLMs or shipping AI-powered features falls somewhere under it. Using my open-source project as a live case study, I'll show how to classify an AI tool under the Act's risk categories and what obligations apply when you build on top of GPT, Claude, or Gemini. We'll map everyday engineering practices to actual compliance requirements: logging as record-keeping, code review as human oversight, documentation as technical evidence. You'll leave knowing what applies to your project today, what's coming in 2026–2027, and a practical checklist to get ahead of it. Project: https://github.com/aiqualitylab/ai-natural-language-tests
AI Second Brain for Modern Project Management
In this talk, I’ll share how I’m building an AI Second Brain for modern project management. The idea is to bring together project decisions, context, notes, and operational knowledge in one structured system that can help teams find information, connect the dots, and prepare better decisions. I’ll explain the WIKI-LLM approach and how it turns a traditional knowledge base into something more active and useful in everyday work. I’ll also share my approach and practical principles that make such a system reliable: clear structure, source tracking, human ownership, and regular maintenance.