Tag feed

#rag

1,793 posts68 followers

Explore Hashnode

Alternatives

Trending tags this week

ASAbto Softwareabto-software.hashnode.dev14h ago · 5 min read

Why Most RAG Projects Stall Before They Reach Production

Retrieval-augmented generation looks simple on a whiteboard. Connect a vector database to an LLM, retrieve relevant chunks, generate an answer. In practice, teams that build this way end up with a sys

1

K

SPSahil Patelsahil-genai-notes.hashnode.dev20h ago · 13 min read

# RAG Isn't Magic: How Retrieval-Augmented Generation Actually Works (and Where It Quietly Breaks)

A few weeks ago, I asked a chatbot a simple question about a product's return policy. It gave me a confident, detailed, beautifully worded answer. There was just one problem: none of it was true. The

0

DKDamir Karimovblog.damir-karimov.com13h ago · 17 min read

Why Most RAG Systems Fail in Production: The Hidden Architecture Problems Behind AI Search

Building a reliable RAG system is not about connecting an LLM to a vector database. It requires a complete architecture covering data ingestion, retrieval, ranking, evaluation, and production operatio

0

PMpriyanshu maitycourtroom.hashnode.dev13h ago · 2 min read

How to Build an AI Courtroom with AI Judges, AI Lawyers, and RAG

Most legal AI tools focus on summarising documents or answering queries. While they’re useful, they rarely explain how a decision is reached or consider multiple perspectives before producing an answe

0

AKAntony Kanaguantonykanagu.hashnode.dev20h ago · 11 min read

OKF: A Knowledge Base That Gets Smarter Every Time You Ask

In April, Andrej Karpathy published a GitHub gist he called an idea file: llm-wiki.md, a pattern for letting an LLM agent build and maintain a knowledge base as plain, interlinked markdown. I read it

0

NLNagesh Lakshmana Gowdanageshtech.hashnode.dev1d ago · 11 min read

Designing a Production-Grade Enterprise RAG System: From Upload to Answer

Part 1 of a 2-part series. → Part 2: The 24 Edge Cases That Break Production RAG Systems — what happens after you ship it. Most RAG tutorials show you a 20-line demo: load a PDF, embed it, ask a quest

0

NLNagesh Lakshmana Gowdanageshtech.hashnode.dev1d ago · 13 min read

The 24 Edge Cases That Break Production RAG Systems

Part 2 of a series on enterprise RAG. In Part 1: From Upload to Answer, I walked through the full architecture — the document management system, the retrieval pipeline, and the permission model. This

0

SSSameer Selokarblogbysameer.hashnode.dev1d ago · 10 min read

Retrieval-Augmented Generation (RAG): What It Is, How It Works, and Why It Sometimes Fails?

Most people think connecting an AI model to your documents magically makes it accurate. The reality is more nuanced. Retrieval-Augmented Generation (RAG) significantly improves an LLM's ability to ans

0

JJacksonrecalso.hashnode.dev1d ago · 8 min read

Hybrid search on-device: why we index every filename with SQLite FTS5 and embed only what matters

The bug report that killed our "just use embeddings" plan A user knows the file exists. They typed its name. They can picture the icon. And the semantic search we shipped confidently returned four the

0

PBPranav Bakalepranavbakale.hashnode.dev1d ago · 6 min read

Demystifying RAG: Why Giving AI an "Open-Book Exam" Changes Everything

Imagine you have a brilliant friend who has read almost every book, article, and website in the world up until a year ago. If you ask them to explain quantum physics or write a poem, they’re amazing.

0

#rag

Search Hashnode

#rag

Explore Hashnode

Trending tags this week

Why Most RAG Projects Stall Before They Reach Production

# RAG Isn't Magic: How Retrieval-Augmented Generation Actually Works (and Where It Quietly Breaks)

Why Most RAG Systems Fail in Production: The Hidden Architecture Problems Behind AI Search

How to Build an AI Courtroom with AI Judges, AI Lawyers, and RAG

OKF: A Knowledge Base That Gets Smarter Every Time You Ask

Designing a Production-Grade Enterprise RAG System: From Upload to Answer

The 24 Edge Cases That Break Production RAG Systems

Retrieval-Augmented Generation (RAG): What It Is, How It Works, and Why It Sometimes Fails?

Hybrid search on-device: why we index every filename with SQLite FTS5 and embed only what matters

Demystifying RAG: Why Giving AI an "Open-Book Exam" Changes Everything