Welcome! Here you will find code labs, code snippets, reference applications, open source projects, SDKs and infrastructure tools as you build on VAST.
The VAST AI Operating System unifies storage, database, and compute to transform data into action. The software platform is built from ground up to manage and process massive amounts of unstructured data (files, objects, tables, vectors) at scale:
- Cosmos Labs: Hands-on labs to learn VAST through storage monitoring, metadata, and pipeline scenarios
- DataEngine Pipelines: Collection of reference pipelines that run on VAST DataEngine
- Code Snippets: Collection of useful code snippets to build on VAST
- Video Search and Summary: Build video ingestion and retrieval pipelines. Repo demonstrates how to use data pipeline and vector database for AI applications
- Document Research Assistant: Build document ingestion and retrieval. Repo demonstrates how to use data pipeline and vector database for document based AI applications
- Python SDK: Python SDK for VAST Management System (VMS)
- Go SDK: Go SDK for VAST Management System (VMS)
- VAST Management System MCP: MCP server for AI assistants to monitor and manage VAST clusters
- DataEngine CLI: CLI to manage VAST DataEngine functions, pipelines, triggers, and compute resources
- Database SDK: Python SDK for VAST Database and Catalog via PyArrow-compatible operations
- Orbit: Complete data science toolkit for VAST Database. Scikit-learn-like API with SQL/Python flexibility.
- Database Connectors: VAST Connector for third-party query engines (Trino and Spark)
- Arrow Database Connectivity: Arrow database driver to query VAST Database
- Ansible: Ansible modules to automate VAST infrastructure with idempotent, declarative modules for views, policies, authentication, and networking
- Terraform: Terraform provider to manage VAST infrastructure
- Kubernetes CSI driver: Kubernetes CSI driver to integrate VAST storage with container workloads