Agentic Coding Workshop 006: From Autoresearch to Evals

From Autoresearch to Evals: Shipping AI That Actually Works The way we write code is fundamentally changing. AI coding tools have moved beyond autocomplete—they're now capable collaborators that are reshaping how founding engineers work. Join us for an evening with two builders working at the frontier of applied AI. Niko Lazaris (Obvious) will walk through an autoresearch loop workflow, and a member of the Fractional AI team will dig into evals—how they're used to verify that systems behave as intended in real production environments. Moderated by Arjun Kannan, the session runs as a roundtable: presentations are the jumping-off point, but the real value is in the room. Expect to discuss, contribute, and compare notes throughout—then map out what you're going to try next. About Your Moderator Arjun Kannan , Co-Founder at Residesk, which helps real estate owners make sense of what residents stay and pay for by texting with millions of renters and understanding what they're actually saying. Previously, he was CTO and CPO at Climb Credit, where he built ML-driven financing products that…