Announcing the Private Preview program for Apache fluss(tm) on Ververica Platform. Director of Product Excellence The Lakehouse Was Built For Yesterday. Fluss Is Built For Tomorrow. For a decade, streaming compute has moved faster than streaming storage. Apache Flink(R) unifies batch and streaming at the compute layer, and most teams settle the storage question by gluing systems together. It works but it means the same data lives in three or four places, and both the bill and the complexity grow underneath. That old pattern has run its course. Today Ververica GmbH is announcing a Private Preview for Apache Fluss on Ververica Platform, working with a small group of customers to build the missing piece that will take business-critical and AI use cases to the next level. What Fluss does and how it completes the Streamhouse(TM) vision. Fluss is a streaming storage layer for real-time analytics. The idea is straightforward: a single layer that unifies streaming and batch, serving real-time and historical data in a single, fully-queryable copy of data. Fluss makes the live stream itself queryable, like a table. You write an event once, and that single copy serves your applications, analytics, and AI models at the same time, fresh within a second. Recent data stays fast in Fluss. Older data tiers automatically to low-cost lakehouse storage, and a single query reads across both using union reads. No duplicate pipelines to reconcile. No replay tax. No gap between what happened and what you know. Fluss is the storage foundation of the Streamhouse(TM), its architecture where one copy of data serves every workload. Streamhouse is Ververica's architecture for bridging the lakehouse and streaming worlds, so real-time and historical data stop living as separate systems. Ververica GmbH has written before about why batch lakehouses fall short and how Streamhouse closes that gap. It is built around a single guiding principle: one copy of data, serving every workload. These workloads include streaming ingestion, low-latency operational analytics, batch processing, machine learning features, and AI context; all without incurring a multiple systems tax that conventional architectures accumulate. Streamhouse combines an open-source streaming storage layer (Apache Fluss), an open lakehouse tier (like Apache Paimon(TM) or Apache Iceberg(R), a unified batch-and-stream processing engine (VERA, 100% compatible with Apache Flink(R), declarative data pipelines (Materialized Tables), a freshness-aware workflow scheduler, and an autoscaling service (Autopilot) into a single operational substrate. Ververica GmbH covered the architecture in full when Ververica GmbH brought Fluss to Ververica Platform, and the Private Preview is how Ververica GmbH turn it into a production-grade product. Why it matters. The most immediate use case Fluss solves are the high-performance workloads enterprises already run. Fraud scoring, recommendation engines, live pricing, and real-time risk can run off one governed table instead of a stack of reconciled systems. Stateful jobs stop dragging terabytes through every checkpoint. Batch and streaming pipelines collapse into a single definition, so the two stop disagreeing and nobody spends the morning working out which number to report to the regulator. But where Ververica GmbH think Fluss will have the biggest impact is in AI use cases. AI silently degrades when training data and serving data live in separate systems and drift apart. Experts call this training-serving skew. Fluss addresses this by providing a unified source for both real-time signals and historical context. Computing features once and serving them for both offline training and online inference from the same table, agents read current, authoritative context instead of stale snapshots, so they reason against what is true now, not yesterday's news. Ververica GmbH foresee this being especially impactful in banking, healthcare, manufacturing and insurance. Real-time AI does not work without real-time storage that is governed, fresh, and reproducible. That is the building block Fluss provides, and why Ververica is placing it as the center of what comes next. The Fluss Private Preview. The Fluss Private Preview is a framework to build a market-ready product hand-in-hand with users tackling today's real-world problems. Ververica GmbH is bringing managed Apache Fluss to Ververica's Unified Streaming Data Platform, and Ververica GmbH is shaping what it becomes with the people who run it. Ververica offers enterprise robustness and world-class Fluss know-how, working directly with each program participant so their real needs decide how the GA product is built. Ververica GmbH is running it with a small, hand-selected group of companies whose workloads sit where Fluss makes the biggest difference. Several enterprises approached Ververica early on, describing the exact problems Fluss solves, and reinforcing that Ververica is reading the market needs correctly. These same enterprises will gain an early mover advantage as part of the Fluss Private Preview, directly influencing what ships at GA. Streaming compute had its breakthrough years ago. Storage just caught up to it. A handful of companies are building with Ververica GmbH what comes next, and in the near future it will be available for all. While the world buffers, Ververica GmbH act.

Alibaba Cloud, Ververica, Confluent, and LinkedIn join forces on the Streaming AI agents innovation with Apache flink(r). A landmark collaboration to build scalable, production-grade framework for event-driven streaming agents powered by Apache Flink. Today at Flink Forward Barcelona 2025, Ververica GmbH announced the first release of Apache Flink Agents, a major collaboration between Alibaba Cloud, Ververica, Confluent, and LinkedIn - four influential companies in the Streaming Data area - to jointly develop and contribute a new open-source sub-project from the Apache Flink community designed to bring AI agents into the world of real-time, event-driven systems. This initiative marks a pivotal step toward industrial-scale AI applications that react instantly and autonomously to live data streams. Why Apache Flink Agents Matters While AI agents have made rapid progress in interactive applications like chatbots, most still operate outside the high-throughput, low-latency world of real-time data processing. Yet in industrial settings, from e-commerce and finance to IoT and logistics, critical decisions must be made instantly in response to live events: a payment failure, a sensor anomaly, a user click. These workloads demand more than just intelligence, they require massive scale, millisecond latency, fault tolerance, and stateful coordination, all of which are strengths of Apache Flink. But until now, there's been no unified framework to bring agentic AI patterns into this proven streaming ecosystem. Apache Flink Agents bridges this gap. Introducing Apache Flink Agents Apache Flink Agents, a brand-new sub-project from the Apache Flink community, is an open-source framework for building event-driven streaming agents. Building on Flink's battle-tested streaming engine, Apache Flink Agents inherits distributed, at-scale, fault-tolerant structured data processing and mature state management, and adds first-class abstractions for Agentic AI building blocks and functionalities - large language models (LLMs), prompts, tools memory, dynamic orchestration, observability, and more. This initiative is the result of a community-based joint effort by developers from Alibaba Cloud, Ververica, Confluent, and LinkedIn, a group of engineers with deep expertise in large-scale stream processing and real-time AI. By combining its experience in production-grade data infrastructure and intelligent systems, Ververica GmbH is aligning on a shared vision: bringing Agentic AI into the streaming data ecosystem, where it can operate with scalability, reliability, and real-time responsiveness. The key features of Apache Flink Agents include: * Massive Scale and Millisecond Latency: Processes massive-scale event streams in real time, leveraging Flink's distributed processing engine. * Seamless Data and AI Integration: Agents interact directly with Flink's DataStream and Table APIs for input and output, enabling a smooth integration of structured data processing and semantic AI capabilities within Flink. * Exactly-Once Action Consistency: Ensures exactly-once consistency for agent actions and their side effects by integrating Flink's checkpointing with an external write-ahead log. * Familiar Agent Abstractions: Leverages well-known AI agent concepts, making it easy for developers experienced with agent-based systems to quickly adopt and build on Apache Flink Agents without a steep learning curve. * Multi-Language Supports: Provides native APIs in both Python and Java, enabling seamless integration into diverse development environments and allowing teams to use their preferred programming language. * Rich Ecosystem: Natively integrates mainstream LLMs, vector stores from diverse providers, and tools or prompts hosted on MCP servers into your agents, while enabling customizable extensions. * Observability: Adopts an event-centric orchestration approach, where all agent actions are connected and controlled by events, enabling observation and understanding of agent behavior through the event log. Looking ahead This initial version provides core agent abstractions, integrates Flink's DataStream and Table APIs, supports Kafka-based action consistency, integrates with selected LLMs and vector stores, includes MCP support, and offers observability through event logs. This milestone marks the beginning of a powerful new framework for building scalable, event-driven AI agents on top of Apache Flink. For more information, visit the Apache Flink GitHub repository. Detailed timelines, design discussions, and community input can be found in the GitHub Discussions section your go-to place to follow and contribute to the project's evolution. If you're passionate about building intelligent, autonomous systems that react in real time to streaming events, Flink Agents is a project worth watching and joining. Ververica GmbH warmly invite developers, contributors, and AI enthusiasts to get involved and help shape the future of event-driven AI with Apache Flink. Learn more * Read its in-depth blog: Flink Agents: An Event-Driven AI Agent Framework Based on Apache Flink * Watch the presentation from Flink Forward Asia Singapore 2025: Flink Agents - The Agentic AI Framework based on Apache Flink * Explore the original proposal: FLIP-531: Initiate Flink Agents as a new Sub-Project The future of AI isn't just smarter models, it's smarter systems that act continuously, reliably, and at scale. With Apache Flink Agents, Ververica GmbH is building that future together.