StreamNative Announces Ursa Engine GA on AWS and UniLink Migration Tool, Slashing Data Streaming Costs 10x for AI-Ready Lakehouses

SAN FRANCISCO, March 18, 2025 /PRNewswire/ -- StreamNative, the cloud-native data streaming company, today announced the General Availability (GA) of Ursa Engine for Bring Your Own Cloud (BYOC) on Amazon Web Services (AWS), which natively integrates with Snowflake Open Catalog, Databricks Unity Catalog, and Amazon S3 Tables to seamlessly stream real-time data into AI-ready data lakehouses. Along with Ursa Engine GA, StreamNative announces UniLink (Universal Linking) public preview to help customers migrate from legacy data streaming platforms to Ursa. These offerings drastically reduce infrastructure costs, simplify Apache Kafka® migrations, and unify governance across streaming and batch workloads—making real-time data a foundational asset for AI and analytics.

The General Availability of Ursa Engine – a Kafka-Compatible, Lakehouse-Native Data Streaming Engine – on AWS

For years, data streaming has been limited to niche, resource-intensive applications, such as predictive maintenance, customer behavior analysis and real-time dashboards due to the complexity and expense of traditional streaming platforms, like Kafka. While some vendors have positioned Kafka as a solution for large-scale data ingestion (ETL), its high costs—especially during data surges—have hindered adoption. Ursa Engine breaks this cost barrier, making real-time streaming viable for AI, analytics, and large-scale lakehouse ingestion.

Kafka's traditional leader-based architecture introduces significant infrastructure costs, complex partition management, and expensive inter-AZ replication. Ursa Engine eliminates these inefficiencies with a leaderless architecture that removes the need for leader elections, reducing operational complexity and inter-AZ traffic. It also adopts a lakehouse-native storage design, enabling organizations to store data directly in object storage using open table formats such as Apache Iceberg™ and Delta Lake. This architectural shift drastically reduces infrastructure costs by a factor of ten, making real-time data streaming a practical and cost-effective solution for AI pipelines, analytics, and large-scale data ingestion.

"As a longtime StreamNative customer running its classic Apache Pulsar™ engine in production for our ultra-low-latency workloads, I couldn't be more excited about the new Ursa Engine GA. It opens up entirely new possibilities for data-intensive, latency-relaxed workloads, all while simplifying our enterprise data architecture. You don't have to worry about cost anymore—just pump all your streaming data into Ursa. Our evaluation shows it to be 10 times more cost-efficient than other Kafka solutions. Everything is seamlessly written to object storage, automatically compacted into Iceberg tables, and made immediately available for our data teams using Snowflake. Ursa Engine delivers the performance, cost-efficiency, and scalability we need to stay ahead in our industry," said Christos A, Enterprise Architect at a F500 company.

Additionally, Ursa Engine natively integrates with modern data catalogs — including Snowflake Open Catalog, Databricks Unity Catalog, and AWS S3 Tables — to provide organizations with end-to-end data governance for both streaming and batch workloads. This eliminates the need to set up bespoke integrations to connect Kafka with Lakehouse via connectors or custom integrations, putting streaming data and lakehouse data to understand the same governance model.  This ensures consistent security, access control, and lineage tracking, making real-time data immediately usable for AI training and analytics without additional ETL overhead.

Christian Kleinerman, EVP of Product at Snowflake, added, "Real-time data ingestion is a critical component of modern AI and analytics. Ursa Engine's integration with Snowflake Open Catalog and Apache Iceberg will allow businesses to stream data into their lakehouses more seamlessly, reducing costs while maintaining and prioritizing unified governance, interoperability, and performance. We're excited to see StreamNative bring this powerful solution to the market."

Universal Linking enters public preview: Full-Fidelity Kafka-to-Ursa Migration Without Compromise

To further simplify migration from Kafka, StreamNative introduces UniLink, a seamless Kafka replication tool that allows organizations to replicate Kafka topics and metadata to Ursa Engine without downtime. UniLink provides a seamless data migration tool to accelerate customers to transition from a legacy leader-based architecture towards Ursa and accelerate their journey to make streaming data ready for AI and analytics. By leveraging smart zone-aware reads and direct object storage integration, UniLink eliminates broker bottlenecks and minimizes expensive cross-AZ data transfers. This approach enables seamless, high-performance replication while significantly reducing infrastructure overhead. With UniLink, organizations can scale their data streaming operations effortlessly—replicating more while spending less, without compromise. UniLink enables businesses to convert Kafka topics directly into Iceberg or Delta tables, ensuring full compatibility with modern data lakehouses while significantly reducing operational costs.

Sijie Guo, CEO & Co-Founder, StreamNative, said,  "For too long, high costs have kept real-time data streaming out of reach for many organizations. With today's announcements and our catalog integrations, we're making data streaming more affordable, scalable, and AI-ready—without disrupting existing Kafka workloads."

Experience New Ursa and explore StreamNative's innovative solutions with a $200 credit at console.streamnative.cloud.

About StreamNative
StreamNative, the leader in real-time data streaming, offers a high-performance, cost-efficient data streaming platform powered by Ursa Engine, supporting mission-critical operational business applications, AI, and analytics workloads. Built on a leaderless, lakehouse-native architecture, StreamNative eliminates the inefficiencies of traditional systems like Kafka, enabling enterprises to operate at 5% of the cost while unifying streaming and batch data in open formats such as Apache Iceberg and Delta Lake. Trusted by global enterprises and fast-growing unicorns, StreamNative offers unmatched flexibility with Serverless, Dedicated, BYOC (Bring Your Own Cloud), and Private Cloud deployment options—all backed by a 99.95% SLA and 24/7 expert support. Recognized as a Leader in the 2024 GigaOm Radar Report for Streaming Data Platforms, StreamNative ensures real-time data flows with zero operational overhead, empowering organizations to transform raw data into AI-ready insights at unprecedented speed and scale.

SOURCE StreamNative