3d ago · 9 min read · The motivation of the article is use AI with my financial statements as context and ask questions around those. In order to build everything locally had to use RAG approach. The statements in this cas
Join discussionMay 23 · 13 min read · Recently, I’ve been exploring a lot about RAG systems and local LLMs, not from a research perspective, but more from a curiosity perspective as a developer who genuinely enjoys building things and und
Join discussion
May 15 · 16 min read · This is Part 10 of the AI Engineering with TypeScript series. Prerequisites: Part 2 — MCP Fundamentals · Part 3 — AI Agent · Part 5 — Production MCP Server Stack: Node.js 20+ · TypeScript 5.x · OpenAI
Join discussion
Apr 17 · 12 min read · RAG now drives 51% of enterprise AI implementations — up from 31% just a year ago. Behind every one of those pipelines is a vector database making millisecond-level similarity decisions across millions or billions of embeddings. Choosing the wrong on...
Join discussion
Apr 14 · 17 min read · 📂 Explore the Source If you want to dive straight into the implementation, you can explore the repositories below: Srot Core (github.com/pathak404/srot-core): The central intelligence pipeline, Fast
Join discussion
Apr 7 · 17 min read · A client came to me in early 2024 with a fully built AI agent that kept timing out in production. The retrieval step took 800ms on average and occasionally spiked past two seconds. Their users were abandoning queries. The "AI is too slow" complaint w...
Join discussion
Mar 30 · 26 min read · If you’ve ever built a sensor pipeline before, you know how it usually ends. Data flows in, you set some thresholds, draw a pretty chart, and call it a day. And honestly? That’s fine. Until it isn’t.
Join discussionMar 13 · 16 min read · I was building a venue discovery feature. The idea was simple: a user describes what they are looking for in natural language, and the app surfaces the best matches from a catalogue of thousands of pl
Join discussionMar 3 · 15 min read · Vector Databases for RAG: Qdrant, Weaviate, and Milvus Compared Large language models are powerful but they hallucinate, go stale, and have no knowledge of your private data. Retrieval-augmented generation (RAG) fixes this by fetching relevant contex...
Join discussionFeb 21 · 4 min read · Building a Private AI Memory System: Local Mem0 with Qdrant + Ollama Ever wanted your AI assistant to remember things... but without sending all that data to the cloud? I've been using Mem0 for my age
Join discussion
