cocoindex-v1

What problem does it solve?

CocoIndex v1 provides a Python-native framework for building incremental data processing pipelines by declaratively specifying target states and letting CocoIndex manage incremental updates and change detection.

Core Features & Use Cases

• Incremental processing: re-run only changed data and propagate updates to targets like databases and vector stores.
• Declarative targets: declare desired end state; CocoIndex handles creation, update, and deletion automatically.
• Flexible data types: supports any Python types with optional vector embeddings and database integrations.
• Memoization: avoid recomputations for unchanged inputs or code.
• Multi-database/FS targets: integrates with PostgreSQL, SQLite, LanceDB, Qdrant, and LocalFS for pipelines, embeddings, graphs, and ETL.

Quick Start

Start building with CocoIndex v1 by initializing a project, installing dependencies, and running the update command:

• Run cocoindex init my-project to create a new CocoIndex v1 project.
• Install dependencies with pip install -e .
• Run your app with cocoindex update main.py to execute the pipeline.