risingwavelabs/risingwave

Event streaming platform for agentic AI. Continuously ingest, transform, and serve event streams in real time, at scale.

What it solves

RisingWave is designed to eliminate the complexity of traditional event streaming stacks. Instead of chaining together multiple tools like Debezium, Kafka, and Flink for data ingestion, processing, and serving, it provides a single system that ensures data is always fresh and queryable at low latency for AI agents and real-time applications.

How it works

RisingWave continuously ingests data from webhooks, database changes (CDC), event streams (Kafka, Pulsar, Kinesis), and historical batch data from S3. It uses incremental computation to update results only when upstream data changes, maintaining end-to-end freshness under 100ms.

Results are served via an internal row store for low-latency (10-20ms) SQL queries, while long-term storage and analytical access are handled through native integration with Apache Iceberg tables, which are managed automatically by the system.

Who it’s for

It is built for developers building agentic AI and real-time applications that require a unified pipeline for streaming data, feature stores, and live dashboards.

Highlights

  • Unified SQL Interface: Joins streams and tables using standard SQL via the PostgreSQL wire protocol.
  • Incremental Computation: Updates materialized views without full recomputation, ensuring high freshness.
  • Cost-Efficient Storage: Stores internal state and tables in object storage (S3) to reduce costs compared to RAM.
  • Agent-Ready: Includes an MCP server, CLI, and Skills to allow AI agents to query and operate the system directly.
  • Open Format Integration: Natively manages Apache Iceberg tables for durable, open-format storage.

Related

  • Project
    smaramwbc/statewave

    Statewave is an open‑source, self‑hosted memory runtime for LLM agents. It records raw events (episodes), compiles them into typed memories (via heuristic or LLM compilers), and serves deterministic, token‑bounded context bundles with full provenance. Built on PostgreSQL + pgvector, it offers policy‑driven labeling, multi‑tenant isolation, and SDKs for Python and TypeScript.

  • Project
    apache/flink-agents

    An agentic AI framework based on Apache Flink that allows developers to build AI agents capable of real-time stream processing.

  • Project
    airweave-ai/airweave

    An open-source context retrieval layer for AI agents and RAG systems that syncs data from 50+ integrations into a unified, LLM-friendly search interface.

  • Project
    AgentEra/Agently

    An AI application runtime framework that provides structured outputs, observable actions, and signal-driven workflows to build reliable, production-ready AI service backends.

  • Project
    Atmosphere/atmosphere

    A real-time engine for AI agents on the JVM that provides streaming transports, enterprise governance, and durable session management for production-grade agent services.