Why it matters
Hands-on use of a production agentic coding tool clarifies where AI assistance becomes autonomous execution — and where that boundary matters for software architecture decisions.
The tokenmaxxing angle
Claude Code spends tokens on every codebase read, plan, edit, and test loop. A 90-minute build sprint makes the token cost of agentic iteration concrete and teaches builders where to constrain scope to manage spend.
From the organizers
Hosted by NYC Applied AI Builders. Format: 15-minute Claude Code walkthrough, then a 90-minute solo build. Attendees bring an idea or pick from provided suggestions; builds run live on personal laptops.