<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Writing · Hussam Elnemr</title><description>Notes on the judgment calls behind the systems: what I chose, what I didn’t, and why.</description><link>https://lilhuss26.me/</link><item><title>Denormalized pgvector on the retrieval hot path</title><link>https://lilhuss26.me/writing/denormalized-pgvector-retrieval-hot-path/</link><guid isPermaLink="true">https://lilhuss26.me/writing/denormalized-pgvector-retrieval-hot-path/</guid><description>Keep the source of truth normalized. Give retrieval a flat copy of its own, so the hot path never pays for a join.</description><pubDate>Sun, 22 Nov 2026 00:00:00 GMT</pubDate></item><item><title>Hybrid retrieval with BM25 and RRF</title><link>https://lilhuss26.me/writing/hybrid-retrieval-bm25-rrf/</link><guid isPermaLink="true">https://lilhuss26.me/writing/hybrid-retrieval-bm25-rrf/</guid><description>Why keyword search still earns its place next to vectors when every answer has to cite a legal article.</description><pubDate>Sun, 22 Nov 2026 00:00:00 GMT</pubDate></item><item><title>Why I pick LangChain over LangGraph for multi-stage agents</title><link>https://lilhuss26.me/writing/why-i-pick-langchain-over-langgraph/</link><guid isPermaLink="true">https://lilhuss26.me/writing/why-i-pick-langchain-over-langgraph/</guid><description>A pipeline with many stages is not a graph problem. What decides it is how many decisions the agent makes, and how much state has to survive between them.</description><pubDate>Sun, 22 Nov 2026 00:00:00 GMT</pubDate></item></channel></rss>