Full-Stack AI Engineering from AI-Native Coding to Knowledge, Models, Infrastructure & OSS
summary
Conference lays out concrete claims and failures: modern dev tools like copilots and repet can produce clickable code that breaks test workflows when dynamic IDE UI hides test tabs, forcing manual debugging. It treats LLMs as a runtime that compresses knowledge but shows high indeterminism, so teams extract, validate and enrich facts into knowledge graphs via an agentic architecture with AI memory. It flags inference as the scaling cost — models served millions of times require more GPUs and smarter AI data infrastructure, and pushes interoperable open source stacks.
AI-written summary · about
chapters (5)
- 0:00Introduction to the conference
- 1:45AI-native coding practices
- 5:20Knowledge data and models
- 8:06AI data infrastructure
- 11:00Open source AI ecosystems
related talks
09mLessons Learned in AI Engineering
47mEmbedding AI: Don’t Burn Your Fingers in the Heat of the AI revolution
40mOpening Keynote: AI & Platform Teams
24mSoftware 3.0 and the AI Engineering Landscape
47mAI and the Future of Software Development
15mKeynote: The Common Pathway of AI and Software
31mModern Architecture for AI-Augmented Development Teams
46mFullstack AI: The Building Blocks for Production Ready AI-Powered Software