Why it matters
Quantitative assessment frameworks for AI coding agents — like the Agent GPA approach shown here — are exactly the tooling needed to decide which model or tier justifies its cost on real SQL and Python workloads.
The tokenmaxxing angle
The Agent GPA framework for evaluating coding agents by output quality is a direct input to cost-per-outcome analysis. If you can score agent performance numerically, you can route to cheaper models that meet the quality bar.
From the organizers
Presenter Eduardo Gonzalez is an AI/ML Architect at Snowflake's Applied Field Engineering division; event held at Nomadworks, 240 West 40th Street, New York, with complimentary pizza provided.