Datris

•
14
•

Orchestrates comprehensive data pipelines for ingestion, validation, transformation, storage, and retrieval, designed for both AI agents and human developers.

Datris is the first AI Agent-Native Data Platform, offering a comprehensive open-source solution for managing the entire data lifecycle. It allows users to ingest, validate, transform, store, and retrieve data using simple JSON configurations, AI instructions, JavaScript functions, or REST endpoints. With a built-in MCP server, it empowers AI agents to natively interact with data pipelines, perform semantic searches, and query various databases through natural language. Designed for self-hosting across any cloud or on-premise environment, Datris ensures no vendor lock-in and provides intelligence at every stage of data engineering.

Características Principales

010 GitHub stars
02Native MCP (Model Context Protocol) server for AI agent integration and interaction via natural language
03AI-powered data quality, transformations, schema generation, profiling, and error explanation
04Integrated RAG pipeline with multiple vector databases, chunking strategies, and embedding providers
05Configuration-driven data pipelines defined via JSON, AI instructions, JavaScript, or REST endpoints
06Parallel data output to object stores, SQL/NoSQL databases, Kafka, ActiveMQ, and various vector databases

Casos de Uso

01Developing and deploying AI Agent-native data pipelines for real-time data interaction and processing
02Building Retrieval-Augmented Generation (RAG) systems with integrated document extraction, chunking, and vector database upserts
03Automating data quality checks, transformations, and schema generation across diverse data sources using natural language AI