AI-agent testing often starts with an expensive loop: call the agent, send its answer to another...
How to Test AI Agents Without Calling More LLMs
Related
Corrective RAG for billing: the bug is not retrieval, it's the model narrating correct numbers wrong
A production billing assistant, the CRAG knowledge-refinement step done without a second model, and a grounding verifier that rejects any number the evidence cannot support.
A Lei de Conway (DRAFT com Mermaid): por que a sua arquitetura é o espelho da sua organização
Melvin Conway publicou em 1968 uma observação que arquitetos de software ignoram por décadas,...
How to Test AI Agents Without Calling More LLMs
AI-agent testing often starts with an expensive loop: call the agent, send its answer to another...