Mar 26 · 2 min read · The Qdrant MCP server takes a deliberately minimalist approach to AI-native vector search. Just 2 tools — store and find — but it's the only vector DB MCP server supporting all three transport protocols. At a glance: 1,300 GitHub stars, 242 forks, v0...
Join discussionMar 10 · 6 min read · I sat down to write the pgvector section of this post—the HNSW index DDL, the reranker batching, the metadata filter shapes—and realized I kept reaching for the wrong file. The query I was proud of wasn’t the vector search. It was dedupeRagChunks. Th...
Join discussion
Mar 10 · 11 min read · I didn’t switch to multi‑vector embeddings because it was trendy. I did it because a single pooled vector kept lying to my search. When you collapse a candidate into one embedding, you’re asking one point in space to simultaneously represent “career ...
Join discussion
Mar 9 · 15 min read · TLDR: Embeddings convert words (and images, users, products) into dense numerical vectors in a geometric space where semantic similarity = geometric proximity. "King - Man + Woman ≈ Queen" is not magic — it is the arithmetic property of well-trained ...
Join discussionMar 9 · 16 min read · TLDR: A vector database stores meaning as numbers so you can search by intent, not exact keywords. That is why "reset my password" can find "account recovery steps" even if the words are different. 📖 Searching by Meaning, Not by Words A standard d...
Join discussionMar 8 · 3 min read · TL;DR Every vector database used by RAG systems, semantic search, and AI-powered recommendations creates a permanent fingerprint of your data through embeddings. Attackers can use cosine similarity queries to extract original documents, identify user...
Join discussionMar 8 · 8 min read · Three months ago I shipped a RAG pipeline that I was genuinely proud of. Semantic search over our internal docs, OpenAI embeddings, Pinecone on the backend. It felt modern. Then someone on our team asked it "what's our parental leave policy?" and it ...
Join discussionMar 8 · 12 min read · TL;DR When you use AI assistants with memory (Claude, ChatGPT with long-term context, Perplexity), your conversations are converted into vector embeddings—high-dimensional mathematical representations stored in databases. These embeddings can be reve...
Join discussionMar 8 · 11 min read · TL;DR Retrieval-augmented generation (RAG) systems are deployed by thousands of companies to let ChatGPT/Claude access private knowledge bases. The problem: vector embeddings are NOT one-way functions. Attackers can reverse-engineer embeddings to rec...
Join discussion
