Agents Anonymous #8: Jev & the New Normal — Who Reviews All This Code?

​Agents write more code than we can review. What happens when judgment becomes the bottleneck?

​Round 8: Jev & the New Normal — Who Reviews All This Code?

​We'll explore Jev, TypeSafe's model for fast, typed probabilistic decisions, alongside the everyday reality of reviewing agent-generated pull requests. Where can automated screening help? What evidence do we need before trusting it? And which decisions should stay with a human?

​Bring your workflows, failures and questions.

​A safe space for builders using agentic coding tools. Claude Code, Codex, Cursor, Copilot, Pi, Aider, Cline or your own setup: whatever you use, you're welcome. Developers, founders and engineers at any experience level can join. Bring what you've actually tried.

​What makes this different: Nothing is recorded or live-streamed. What's said in the room stays in the room. Real hands-on experiences, including what broke. Jeronim opens with the first lightning talk (5–10 minutes), then opens the floor for guests and contributors to showcase their presentations. Guest lightning talks: around 5 minutes, maximum 1–2 slides. Every presenter gets a kefir, as always 😅 Open discussion about what works and what doesn't.

​Format (all times Europe/Berlin): 18:00 — Arrive and get together 18:45 — Jeronim’s opening lightning talk: Jev, the new normal and the review bottleneck (5–10 minutes) 18:50–18:55 onwards — The floor is open for guest and contributor presentations 19:45 — Open discussion: screening, evidence and human judgment 20:30 — Wrap-up and informal conversations 21:00 — End

​Location: STARTPLATZ Köln. Exact location and room details are shared with approved guests. Free to attend; registration requires approval. Tell us about your tools and setup, and whether you'd like to give a lightning talk.

​Background reading: https://typesafe.ai/blog/introducing-system-one-models-and-jev