A place for real developers
The gap between
knowing and building
closes here.
Structured notes on AI, Python, databases, DSA, and software engineering — written the way you'd explain it to a colleague who actually gets it.
"Every post earns its reading time before it goes up. Depth over frequency. Clarity over coverage."
Latest posts
The Retrieval Pipeline: How a Question Becomes the Right Context
Tracing the full path from a raw user query to the exact chunks handed to an LLM — query processing, candidate retrieval, ranking, top-k selection, and metadata filtering working together.
Natural Language ProcessingChunking Strategy for RAG: The Decision That Quietly Determines Retrieval Quality
Why how you split documents matters as much as what embedding model you use — fixed, recursive, semantic, and structure-aware chunking strategies, plus how to actually choose chunk size and overlap.
Natural Language ProcessingData Ingestion for RAG: Turning Messy Real-World Documents Into Usable Text
The unglamorous but decisive first step of any RAG pipeline — parsing PDFs, OCR, HTML, DOCX, databases, and APIs, and why data cleaning quietly determines how good your retrieval can ever be.
Natural Language ProcessingVector Databases Explained: Where Your Embeddings Actually Live
What a vector database really is under the hood — storage architecture, indexing, metadata filtering, and the operational details that separate a demo from a production RAG system.
Natural Language ProcessingApproximate Nearest Neighbor(ANN) Algorithms: The Search Techniques Powering Every Vector Database
Inside the algorithms that make vector search fast at scale — exact vs approximate nearest neighbor search, and how HNSW, IVF, product quantization, ScaNN, and DiskANN each trade accuracy for speed differently.