Blog Posts

4-bit Rotational Quantization

HFresh: Memory-Efficient Vector Search

Build Semantic Search for Legal Documents with Pinecone and GPT-4

How to Implement Retrieval-Augmented Generation (RAG) with Re-Ranking

I Built an Agentic Hybrid RAG System with FAISS and BM25

Building an Agentic Hybrid RAG System with FAISS, BM25, and smolagents

Your AI Feature Doesn't Need Pinecone. It Needs pgvector.

Case Study: Turning Missed Calls into Booked Leads With an AI Receptionist

Web Scraping API for AI Agents and RAG Pipelines

Building a Hybrid RAG System with FAISS, BM25, and Agentic AI

Why Vector Databases Are Critical for RAG Systems

Weaviate 1.39 Release

INTRODUCTION TO RAG (RETRIEVAL AUGMENTED GENERATION)

# Building a Personal Notes Assistant with RAG, Amazon Bedrock, and Pinecone

How to Build an Agentic RAG Pipeline with Real-Time Web Search

Your RAG Pipeline Doesn't Have an Accuracy Problem - It Has an Evaluation Problem

The Overlooked Reason Your RAG Pipeline Keeps Returning Garbage

Your RAG Demo Works Because Someone Picked the Documents

When RAG Says Duplicate but the LLM Disagrees: Building an Adjudication Layer

Dividing your RAG score by retrieval recall overstates your generation quality, and here is by how much

Building Semantic Search: From Embeddings to a Production-Ready Search Engine

Chunking: the most underrated decision in your RAG pipeline

ElastiCache for RAG: How to Store and Query Embeddings in Redis with Node.js and TypeScript

The model had the right evidence. It still got the answer wrong. I ran a controlled 60-question RAG experiment to find out whether retrieval or the reader was the real bottleneck. The results surprised me. When Better Retrieval Doesn't Mean Better Answer