Atlas

Because "trust me bro" isn't an enterprise retrieval strategy.

Atlas carries the context so your LLM stops hallucinating with 100% confidence.

Pipeline & Workbench Modules 8 Interactive Modes

Ingest Feed files

Chunk Chop chop

Embed Math magic

Pipeline Wire DAG

Execute Pull trigger

Eval Receipts

Chat Talk to docs

Settings System config

Document Registry

Parse, index, and manage your local document repository. Ingested files are immediately ready to be chunked and vectorized for RAG pipelines.

Parse, index, and manage your repository.

Workspace Files

0 total documents

Successfully Ingested

0 ready for chunking

Pending Ingestion

0 waiting to parse

Ingestion Progress

Select any file from the sidebar to inspect or ingest individually.

Chunking Laboratory

Benchmark chunkers and preview document splits dynamically.

Chunking Strategy

Document Workspace

First, select an ingested file from the Workspace.
Then select a strategy to start chunking.

Chunking Laboratory

Partition parsed text into optimal segments for vector embedding.

Partition text into chunks.

Total

Chunked

Pending

Recursive Character

Splits text progressively by paragraphs, newlines, and spaces with overlap.

Use Case: Articles, Markdown notes, PDFs.

Code-Aware

Splits source code along functions, classes, and language syntax boundaries.

Use Case: Python, JavaScript, TypeScript, software.

Semantic

Calculates embedding similarity and splits when there is a shift in topic.

Use Case: Research papers, essays, long-form text.

Markdown Structure

Preserves header hierarchy and splits content cleanly by document sections.

Use Case: Technical spec sheets, user guides.

Sentence Window

Extracts single sentences and attaches surrounding context windows.

Use Case: Fact retrieval, precision Q&A.

Fixed Size

Splits text into fixed character or token lengths with uniform overlap.

Use Case: Baseline benchmarks, uniform text files.

Select any file from the sidebar to inspect or test chunks.

3D Vector Embedding Studio

Vectorize chunked text and explore semantic space

1 Select Strategies

Chunking strategies for document.

Select a file first

2 Select Embedders

Choose models to execute.

1 selected

Execution Log

3 Already Embedded

Existing embeddings computed for this file.

No file selected.
Select a file to view completed embeddings.

Embeddings

Loading registry...

3D VECTOR SPACE

Select an embedded file and combination from the sidebar to project vectors into the 3D space.

Node Details

0.000

Content

Metadata

Doc:--

Strat:--

DB ID:--

Embedding Studio

Transform partitioned chunks into dense semantic vector representations.

Create semantic vectors.

Total

Embedded

Pending

BAAI BGE Small

384-dim

BAAI/bge-small-en-v1.5

High-efficiency dense retrieval model optimized for rapid vector indexing and query ranking.

Use Case: Real-time local RAG, fast document search, quick benchmarks.

All-MiniLM-L6

384-dim

all-MiniLM-L6-v2

Fast, lightweight general-purpose sentence transformer with minimal memory and CPU execution.

Use Case: Everyday semantic embeddings, lightweight CPU deployment.

All-MiniLM-L12

384-dim

all-MiniLM-L12-v2

Deeper 12-layer transformer providing superior sentence clustering and nuanced representations.

Use Case: High-accuracy document clustering, nuance-heavy retrieval.

Intfloat E5 Small

384-dim

intfloat/e5-small-v2

Asymmetric retrieval model trained specifically on query-to-passage search matching.

Use Case: Knowledge base Q&A, query-to-passage matching.

Paraphrase MiniLM-L3

384-dim

paraphrase-MiniLM-L3-v2

Ultra-compact 3-layer distilled transformer designed for instant inference and semantic equivalence.

Use Case: Duplicate detection, paraphrasing, low-latency APIs.

Select any document from the sidebar to vectorize chunks.

Execution Pipelines

Select a pipeline to inspect its composition, architecture and target use case.

8 Presets

Pipelines 8

Top-K

Ingest

Doc Indexed

Chunker

Strategy fixed_size

Embedder

Model bge_small

Ready to Execute Pipeline

Select a pipeline above, enter a prompt, and click Run to retrieve grounded context and generate a response.

Model: Groq —

Latency: 0.00s

Chunks:

Tokens: 0 in / 0 out

Execution Runs Evaluation

Search and inspect queries, compare runs, and evaluate model answers.

Execute Tab Runs

Standard RAG & custom pipeline runs

0 runs

Loading recent runs...

Chat Tab Runs

Interactive conversational RAG queries

0 runs

No recent chat runs found.

RAG Combat Arena

Full-spectrum pipeline comparison

Awaiting Ignition — Click “IGNITE ARENA” to evaluate all selected pipelines.

PIPELINE

Standard RAG

TOP-K

MODEL

Default Model

INGEST

Doc Indexed

CHUNKER

Strategy fixed_size

EMBEDDER

Model bge_small

Document Intelligence Assistant

Ask questions, synthesize insights, or query complex data grounded directly in your indexed documents.

Conversations

History 0

No sessions yet

Atlas

Document Registry

Chunking Laboratory

Chunking Strategy

Document Workspace

Chunking Laboratory

Recursive Character

Code-Aware

Semantic

Markdown Structure

Sentence Window

Fixed Size

3D Vector Embedding Studio

1 Select Strategies

2 Select Embedders

3 Already Embedded

3D VECTOR SPACE

Node Details

Embedding Studio

BAAI BGE Small

All-MiniLM-L6

All-MiniLM-L12

Intfloat E5 Small

Paraphrase MiniLM-L3

Execution Pipelines

Create Custom Pipeline

Ready to Execute Pipeline

Execution Runs Evaluation

Execute Tab Runs

Chat Tab Runs

RAG Combat Arena

Document Intelligence Assistant

System Configuration

Project Status

Active LLM Provider

API Keys & Endpoints