"TYPO3 Agencies and AI: An Experience Report"
summary
Over six months punkt.de drops its external copywriter, runs an AI Discovery Day with ~30 agencies, and ships a client-facing chatbot by dropping a JavaScript snippet into the cookie banner. Practical wins include a two-LLM judge pattern for governance and using RAG/Rec workflows that query Nextcloud/DB knowledge sources before hitting the LLM. Sexenia AI helps build chat frontends for legacy Windows apps. Heavy costs and energy make running Llama 31 / Gemini locally impractical; Gaia‑X/OpenGPTX gets €15M but won’t replace pragmatic RAG approaches.
AI-written summary · about
chapters (20)
- 0:00Introduction and speaker intro
- 1:06Why talk about AI at a dev conference
- 3:12AI’s disruptive but boring reality
- 5:21AI Discovery Day in TYPO3 community
- 6:45AI as chance not threat for agencies
- 7:48Customer use cases and upselling potential
- 8:49Supporting colleagues with AI adoption
- 11:22TYPO3 AI extensions and community solutions
- 13:52LLM as judge and workflow automation
- 16:41AI success stories at punkt.de
- 18:28AI use cases beyond TYPO3 community
- 21:20RAG and knowledge sources in AI workflows
- 24:12Advantages of RAG over training LLMs
- 26:24Generating example data with AI
- 27:08AI in coding and IDE integration
- 28:11Open source LLMs and European solutions
- 30:17Gaia-X and open GPT X initiative
- 32:05Current AI tools and regulatory gaps
- 33:11Key takeaways and call to action
- 36:29Q&A session
related talks
22mClient side Web AI Agents: Building smarter user experiences for future agentic ...
43mAI w/o Spaghetti: Clean Architecture in the Age of AI
27mBuilding AI applications: Successes and Lessons
15mThe future of AI is now: Real-life case studies for on client-side AI adoption in web apps
33mAgentic AI – What's Next?
52mCut the Hype and Build Your Own AI Agents
29mClient side Web AI Agents Building smarter user experiences for a future age
27mWill AI (Finally) Replace Developers?