Topic hub
RAG & Vector Search
Master retrieval-augmented generation, vector databases, embeddings, and semantic search systems.
All posts · 49
RAG & VECTOR SEARCH01
LightRAG vs GraphRAG (2026): Updates, Storage, Local ModelsGraphRAG 3.3.0 has a graphrag update command, so rebuild-on-every-change is stale. LightRAG vs GraphRAG on local-model fit, Postgres storage and deletes.
RAG & VECTOR SEARCH02
pg_search vs pg_textsearch: BM25 Hybrid Search in Postgrespg_search vs pg_textsearch: both scored 0.68 nDCG@10 on SciFact, double native ts_rank_cd. Replicas, license and RLS decide which one you ship.
RAG & VECTOR SEARCH03
Docling vs MinerU vs Marker (2026): Code and Model LicensesMarker weights trigger at $5M revenue or $5M funding, MinerU at USD 20M monthly group revenue. Code vs weight licenses for all three, quoted and dated.
RAG & VECTOR SEARCH04
Milvus 3.0 Upgrade From 2.x: What Breaks, What Is Opt-InMilvus v3.0.1 ships targetVecIndexVersion at 8 while the release note tells you to set 10. Five conditions gate the image-only upgrade.
RAG & VECTOR SEARCH05
pgvector 2000 Dimension Limit: Why CREATE INDEX Failspgvector indexes 2000 dimensions while the column type stores 16000. A vector(3072) column loads ten million rows, then refuses an index. Four ways out.
RAG & VECTOR SEARCH06
Qdrant Snapshot Restore Returns an Empty CollectionQdrant's docs say a restore left at the default priority returns an empty collection. The v1.19.0 source shows when priority is consulted at all.
RAG & VECTOR SEARCH07
Text Embeddings Inference vs vLLM: Pick an Embedding ServerTEI's token budget is 16,384 per forward pass and v1.9.0 flipped auto-truncate to true. Four embedding servers, compared by failure mode, not by speed.
RAG & VECTOR SEARCH08
Multilingual RAG Pipeline: One Index or One Per LanguageElasticsearch ships 36 stopword language lists and Polish is not one of them. Three index layouts for a corpus that is multilingual by statute.
RAG & VECTOR SEARCH09
Contextual Retrieval vs Late Chunking: Which One to UseOne controlled head-to-head puts contextual retrieval at 72.4 nDCG@10 against late chunking's 61.0, at 120x the indexing time. Where each one earns its cost.