Entity Resolution: One Thing, Many NamesSam K., Samuel Kim, sam@acme.com. Deciding what's the same thing is harder and more important than it looks.14 min read·internals-decoded.hashnode.dev2d ago
Debugging Agent Systems: Traces, Replays, Root CausesAgent failures raise no errors. The logging, tracing, and replay toolkit that makes them debuggable.14 min read·internals-decoded.hashnode.dev5d ago
Evaluating Agents: Golden Sets, Judges, and Shadow ModeSame input, different outputs. How to measure whether a nondeterministic system actually works, step by step as well as answer by answer.17 min read·internals-decoded.hashnode.devNov 4
Guardrails: Input, Process, OutputPrompt injection, tool authorization, leak prevention: the three-layer safety system every agent needs, and the design that holds when a filter misses.15 min read·internals-decoded.hashnode.devNov 1
Self-Improving Loops: Agents That Learn From OutcomesTrack outcomes, mine the wins, update the prompts, guard every update: the shape of reinforcement learning without the gradients.15 min read·internals-decoded.hashnode.devOct 29
