Skip to main content

Store everything. Retrieve what matters.

Project description

Retrivo

Store everything. Retrieve what matters.

License: MIT Python 3.10+


A RAG framework for intelligent document understanding.

Ingest any document. Build knowledge. Get answers.


What is Retrivo?

Retrivo is a Retrieval-Augmented Generation platform that turns your documents into a queryable knowledge base. It goes beyond simple text search — understanding tables, images, charts, and relationships across your entire document collection.

Core Capabilities

  • Multi-Modal Intelligence — First-class support for text, tables, images, and charts
  • Hybrid Search — BM25 + vector search with reciprocal rank fusion
  • Knowledge Graphs — Automatic entity and relationship extraction across documents
  • Smart Chunking — Structure-aware and semantic boundary detection
  • Advanced Retrieval — HyDE, cross-encoder reranking, MMR diversity
  • 26+ File Formats — PDF, DOCX, XLSX, PPTX, images, EPUB, and more
  • Online & Offline — Works with cloud APIs or fully local models

Architecture

Document → Ingest → Chunk → Embed → Store
                                       ↓
Query → Understand → Search → Rerank → Answer

Tech Stack

Layer Technologies
LLMs OpenAI, Anthropic, Ollama, HuggingFace
Embeddings OpenAI, Cohere, Sentence Transformers
Vector Stores ChromaDB, FAISS
Search BM25, Vector, Hybrid + Cross-Encoder Reranking
Documents pdfplumber, PyMuPDF, camelot, python-docx
API FastAPI
UI Streamlit

Something cool is coming.

Stay tuned.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

retrivo-0.2.0.tar.gz (828.2 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

retrivo-0.2.0-py3-none-any.whl (275.9 kB view details)

Uploaded Python 3

File details

Details for the file retrivo-0.2.0.tar.gz.

File metadata

  • Download URL: retrivo-0.2.0.tar.gz
  • Upload date:
  • Size: 828.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.9.6

File hashes

Hashes for retrivo-0.2.0.tar.gz
Algorithm Hash digest
SHA256 d1adfa22f3283a5ee7a214b7ed80aba4a0b1e54d61be2f3363514bd596824b42
MD5 22e974f093a0fcee51713af9f1d0323c
BLAKE2b-256 5685ae8e74d4f0dfffa77d99874dfb326166c4fe8ff3a9040784b8b60bfab94c

See more details on using hashes here.

File details

Details for the file retrivo-0.2.0-py3-none-any.whl.

File metadata

  • Download URL: retrivo-0.2.0-py3-none-any.whl
  • Upload date:
  • Size: 275.9 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.9.6

File hashes

Hashes for retrivo-0.2.0-py3-none-any.whl
Algorithm Hash digest
SHA256 358d3dc4a23ea993ec7e781a3b600b5c5d3dc8acce9d8e32465c2d1cdb13227d
MD5 11c7e688a202d9283dace22249e4f1ca
BLAKE2b-256 99c511f51ba20366413d0c05639a4c729e67c4ede2bc09c6b4935abc611d7c20

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page