TLDR: StreamNative has launched Ursa, a lakehouse-native streaming solution, and Orca, an Event-Driven Agent Engine, creating a unified platform designed to address critical infrastructure challenges for real-time Artificial Intelligence (AI) at enterprise scale. These innovations aim to drastically reduce streaming costs, eliminate data silos, and provide the necessary runtime for production-ready AI agents.
SAN FRANCISCO – StreamNative, a leader in cloud-native data streaming, announced a significant advancement in real-time Artificial Intelligence (AI) infrastructure at the Data Streaming Summit in San Francisco on September 30, 2025. The company unveiled two new offerings: Ursa, a lakehouse-native streaming engine, and Orca, an Event-Driven Agent Engine, which together form a comprehensive platform to tackle the escalating demands of enterprise AI adoption.
The industry is currently grappling with a ‘Real-Time AI Infrastructure Crisis,’ characterized by three major bottlenecks. Firstly, traditional data streaming architectures incur exponentially scaling costs when confronted with AI workloads, primarily due to inefficient leader-based replication and separate ETL pipelines. These legacy systems struggle to handle the massive data throughput required by AI economically, often leading to prohibitive cost structures at scale. StreamNative claims its new solutions can eliminate up to 90% of these streaming costs.
Secondly, the prevalent disconnection between streaming and analytics systems results in fragmented data silos. This forces AI agents to operate on stale data snapshots rather than leveraging live, real-time context, thereby hindering their ability to make informed and timely decisions. StreamNative’s new platform aims to bridge these data silos, ensuring AI agents have access to the most current information streams.
Lastly, while numerous AI agent frameworks exist for research and demonstration purposes, production environments lack the robust event-driven runtime, persistent memory, and comprehensive observability features essential for autonomous systems to operate reliably at enterprise scale. Orca, the Event-Driven Agent Engine, is specifically designed to provide this missing infrastructure, enabling the deployment of production-ready AI agents.
Ursa, the lakehouse-native streaming engine, is now available across all deployment models, including Serverless, Dedicated, and Bring-Your-Own-Cloud (BYOC) environments. This broad availability ensures flexibility for enterprises regardless of their existing cloud strategy. Orca, the Event-Driven Agent Engine, is currently in private preview, indicating a phased rollout to select customers before wider release.
Also Read:
- GoodData Unveils Next-Generation Full-Stack Data Intelligence Platform with Advanced AI Integration
- Databricks Introduces Data Intelligence for Cybersecurity to Fortify Defenses Against AI-Driven Threats
StreamNative’s announcements position the company as a key player in solving the complex infrastructure challenges that have impeded widespread enterprise AI adoption, offering a unified and cost-effective solution for real-time AI at scale.
