Articles for people who
ship the stack
Rewrites on React, Node.js, TypeScript and AI. A translated body is shown when it exists. If it does not, you see the English text and a note that says so.
Tagged: Retrieval
Three Ways Production RAG Fails: Fragmented Context, Drift and Misuse
Three ways production RAG fails: fragmented context from chunking, drift as the knowledge base grows, and misuse despite good retrieval, with layer-level fixes.
11 min readReranking in RAG: Cross-Encoders vs LLM Judges and Their Latency
How a second-stage reranker fixes candidate ordering in RAG, comparing cross-encoders with LLM judges, latency costs, failure modes and how to evaluate them.
14 min readWhy RAG Exists: Parametric Memory, Dense Retrieval and the 2020 Paper
Why RAG exists: the limits of parametric memory, how BM25 and embeddings led to dense retrieval, and what RAG-Sequence and RAG-Token trade off in the paper.
7 min readDebugging RAG Retrieval: Recall@K, BM25, Hybrid Fusion and Reranking
A small experiment on a RAG pipeline that ranked the right chunk fourth, showing how chunking, BM25, rank fusion and reranking change Recall@3 and MRR.
11 min readSeven RAG Architectures Beyond the Naive Pipeline and When Each Fits
Naive RAG breaks in predictable ways. Seven architectures, from contextual notes to multi-agent and memory retrieval, with their trade-offs and when to use each.
9 min readWhy Reranking Is the Cheapest Big Win in a RAG Pipeline
Retrieve 50, rerank to 5: how cross-encoder reranking fixes ordering in a RAG pipeline, the gains and latency to expect, when to skip it, and what to ship.
8 min readWhy RAG Still Gives Wrong Answers: Hybrid Search, Reranking and Recall@K
Clean chunks don't guarantee good answers. See where vector retrieval fails and how fusion, a cross-encoder reranker, HyDE and Recall@K measurement fix it.
6 min readMultiDocFusion: Hierarchy-Aware Chunking for RAG on Long PDFs
MultiDocFusion keeps section context in every chunk of a long PDF: layout detection, OCR, a header-tree LLM and DFS grouping, for 8-15% better RAG retrieval.
7 min readLayer by Layer: Designing a Real RAG Retrieval Pipeline for AI Agents
Plan a RAG pipeline with SourceDocument and DocumentChunk contracts, chunking, embeddings and ChromaDB, all behind a RetrievalService that keeps the agent stable.
10 min read
About these articles
Request a 24h estimate
Need the same stack in a production operator layer? Send the brief — estimate within 24 hours.