@davidhahn

Topic Suggestion — Designing a Function That Knows What to Recommend Without Magic Numbers

9h ago · 8 min read · Phase 4 gave the study system memory of individual problems — when each one is due for review based on SM-2. Phase 5 asks a different question: zoomed out across all topics, what should I actually be

Streaming Structured Output — Incremental JSON Rendering Without a Parser

1d ago · 5 min read · 04-structured-output solved the reliability problem: use forced tool use to guarantee valid JSON back from the model. But it waited for the complete response before doing anything with it. For a backe

1

Evals — Why a Bad Eval Is Worse Than No Eval

4d ago · 6 min read · The first time I ran evals on my LLM playground, the agent scored lower than basic RAG — 7.5 vs 8.9. The agent is more capable: it has live tool data on top of retrieval, it can answer questions basic

KKen commented

Adding Spaced Repetition to an LLM Study System — SM-2, Schema Design, and a Scoring Problem Worth Solving

5d ago · 6 min read · The first three phases of my study system handled the session loop: generate a problem, grade the solution, log the result. Useful, but with a fundamental gap — there was no memory of what I struggled

Tool Use + RAG — When Retrieval Becomes a Decision

6d ago · 5 min read · In the basic RAG module, retrieval was unconditional. Every query triggered the same pipeline: embed the query, search the vector store, stuff the top chunks into the prompt, generate. The model had n