with Mayug Maniparambil (Research Scientist, Fin) · Ivo Garais (AI Research Engineer, Tenable) · Tiernan Haugh (Product Designer, Nile)
AI Tinkerers Dublin gathers at Fin HQ for a technical show-and-tell on the architectural mechanics of LLMs. The headline talk, “Depth vs. Recovery: How Reinforcement Learning Reshapes the Reasoning Vocabulary of LLMs,” is a 30-minute deep dive from Fin research scientist Mayug Maniparambil, followed by rapid five-minute community demos — raw, live walkthroughs of active builds — and a hands-on Science Fair for one-on-one code feedback. A curated, screened room of engineers, founders and researchers working on agentic AI, RAG, evals and production AI infrastructure. Free; attendance is manually screened.