Why it matters
A recurring 20,000+ member NYC AI group means real practitioner talk on where agentic code actually needs guardrails, not vendor slides.
The tokenmaxxing angle
Talk on a tolerance-based framework for AI code risk touches the same question tokenmaxxing tracks: which agent tasks are cheap to automate loosely vs need tight, costlier oversight.
From the organizers
Confirmed speaker Madona Wambua (Google Developer Expert, CUNY adjunct lecturer) presents "AI: The Tolerance-Based Framework" per the event agenda.