Tag feed

#chunking

21 posts0 followers

Explore Hashnode

Alternatives

Trending tags this week

LALiji Alexlijibuilds.hashnode.devSep 8 · 14 min read

I Tried Six Ways to Get Heading Levels Out of a PDF. Docling Already Had One.

Docling's PDF pipeline returns every heading at level 1 unless you switch on the stage that ranks them. I spent a day finding that stage and tried five other things first. This is the comparison, with

0

LALiji Alexlijibuilds.hashnode.devSep 8 · 15 min read

Layout-Aware and Hierarchical Chunking Are Stages, Not Strategies

Every chunking guide lists the strategies side by side: fixed size, semantic, hierarchical, layout-aware. Pick one. The guides define layout-aware as "respect the visual structure of the page" and hie

1

A

LALiji Alexlijibuilds.hashnode.devSep 8 · 12 min read

Docling Gives Every PDF Heading Level 1. One Flag Fixes It.

Docling reads PDFs well, and it took me a day to notice the one thing it was not doing. Every heading it found came back at the same level. My RAG chunks lost their context, retrieval still looked fin

0

KMKrish Mahyavanshikrizh.hashnode.devJul 11 · 7 min read

Where RAG Fails: Understanding the Limitations of Retrieval-Augmented Generation

Large Language Models (LLMs) are incredibly powerful. Before reading further: If you're new to AI, I recommend first understanding what Large Language Models (LLMs) are and what actually happens afte

2

MR

AKAshaaf Khanashaaf.hashnode.devJul 10 · 10 min read

Behind the Illusion of Perfection: Why Your RAG Pipeline Fails in Production

When Large Language Models first took the world by storm, they felt like magic. But developers quickly ran into a brick wall: these models were stuck in the past, bounded by strict "knowledge cutoffs"

0

GKGaurav Kumargauravbytes.devJul 7 · 8 min read

Chunking Strategies in RAG: A Deep Dive

RAG systems live or die on one decision that happens before any embedding is computed: how the source document gets cut into pieces. This post walks through three chunking mechanisms available in Lang

1

K

VKVamsi Krishnablog.vamsiannamreddy.comJun 3 · 3 min read

Day 13: Why chunk boundaries shape answer quality

Split a document in the wrong place and you can make its answer unfindable. That is chunking: nobody's favourite topic, and a bigger lever on RAG quality than most model decisions. It also gets made e

0

PSParikshit Sharmaparikshitiiitb.hashnode.devApr 22 · 6 min read

RAG Chunking Strategies: What Actually Works in Production (And What Quietly Destroys Retrieval Quality)

Every RAG tutorial shows you fixed-size chunking in five lines of code. Nobody shows you what happens six months later when your retrieval quality has collapsed and you can't figure out why. This arti

0

RSRahul Sehrawatai-zero-to-hero.hashnode.devApr 13 · 11 min read

Chunking Is the Whole Game

The single most common mistake I see in RAG projects is this: a team spends a month tuning their embedding model and another month picking a vector store, and their retrieval quality is still mediocre. They conclude "RAG is hard." They start looking ...

0

GSGeet Sharmageetsharma.hashnode.devJan 26 · 2 min read

Chunking, Embeddings, and Vector Drift: Why Most RAG Systems Quietly Decay

Most Retrieval-Augmented Generation systems do not fail loudly. They decay. Accuracy slowly drops. Answers become vague. Users lose trust. Teams often blame the LLM. In reality, the failure usually lives in the vector layer. Three silent killers domi...

0

#chunking

Search Hashnode

#chunking

Explore Hashnode

Trending tags this week

I Tried Six Ways to Get Heading Levels Out of a PDF. Docling Already Had One.

Layout-Aware and Hierarchical Chunking Are Stages, Not Strategies

Docling Gives Every PDF Heading Level 1. One Flag Fixes It.

Where RAG Fails: Understanding the Limitations of Retrieval-Augmented Generation

Behind the Illusion of Perfection: Why Your RAG Pipeline Fails in Production

Chunking Strategies in RAG: A Deep Dive

Day 13: Why chunk boundaries shape answer quality

RAG Chunking Strategies: What Actually Works in Production (And What Quietly Destroys Retrieval Quality)

Chunking Is the Whole Game

Chunking, Embeddings, and Vector Drift: Why Most RAG Systems Quietly Decay