Miatz playground

Chunk Slicer

Drag one slider, watch search quality rise then crash.

Coming to the Build-LabAIE-202Browser (JS)AI-Concept-Vizdev · data

The Demystify signature

Shows the exact chunk boundary that slices a sentence in half and the retrieval score visibly dropping in the same frame -- the invisible precision/context/cost triangle becomes three linked bars the learner breaks on purpose.

How it works

What goes in, what comes out

What it does

Learner drags chunk-size and overlap sliders over a real document, and the doc re-slices live with colored boundary marks, including the moment a boundary cuts a sentence in half. A fixed test query re-runs against every configuration, with precision/recall bars and a per-query cost readout moving in step with the sliders.

You bring

A source document (preset or pasted) and a fixed test query.

You get

A live re-chunked document view, precision/recall bars, and a cost-per-query delta as sliders move.

You control

Chunk-size slider (50-2000 tokens), overlap-percentage slider, strategy toggle (fixed / sentence-boundary / semantic), test-query picker.

Where it's used

Lab companion inside AIE-202's chunking sub-module.

chunk size/overlapfixed vs semantic chunkingthe precision-context-cost triangle

Optional · watch & try

Go deeper, elsewhere

Hand-picked public explainers and open tools that complement this one — always optional, never required, never graded.

RAG Chunking Strategies Explained: Optimize Document Chunking for Maximum Recall, PrecisionCovers the precision/recall/cost tradeoff behind chunk size Chunking Strategies in RAG: Optimising Data for Advanced AI ResponsesCompares fixed-size vs. semantic chunking approaches Why Your RAG Gives Wrong Answers (And 4 Chunking Strategies to Fix It) | LangChain Text SplittersDiagnoses bad-chunking failure modes with concrete fixes Chunk Visualizer (Hugging Face Space)Paste a document and see exactly how different chunking settings slice it Introduction to Vector Embeddings and Vector SearchFoundational explanation of turning text into vectors for similarity search Embeddings & Vector Databases ExplainedConnects embeddings to how a vector database actually stores and searches them

Concepts click when you open the machinery.

Three labs are already live and free — the same hands-on style this playground brings to its module.

Play the free labs Take the DSAT