#GEPA
5 articles with this tag
OutRival says tokenmaxxing means doing more with fewer tokens
OutRival's Alvin Pane told AI Engineer tokenmaxxing should minimize tokens, using a ceiling test and optimizer to cut a 5,000-token prompt 51% and trade 443 regressions for 2,463 gains.
GEPA squeezes 7x gains from three examples
Lakshya Agrawal showed GEPA beating GRPO with 3 examples by reflecting on traces, then generalizing to Optimize Anything.
AI Evals for Cross-Functional Teams at DoorDash
DoorDash turned AI evals into a cross-functional workflow with self-serve judges and vibe-coded annotation UIs built on API-first primitives.
Samuel Colvin on Optimizing AI Agents in Production
Samuel Colvin, Pydantic CEO, discusses optimizing AI agents in production using GEPA and Logfire's managed variables at AI Engineer Europe.
LLM Evaluators: Beyond Naive Judgments
Mahmoud Malaeb of Argenta discusses the limitations of naive LLM judges and introduces GEPA, an optimization framework for building more accurate LLM evaluators using a data flywheel approach.
