Onehouse builds a fully managed data lakehouse platform that jumpstarts ingestion and data standardization into open data formats and interoperates with major cloud-native services and engines. The company was founded by Vinoth Chandar, creator of the open source Apache Hudi project, and leverages Hudi’s ACID and metadata features to enable real-time upserts, deletes, and low-latency queries on diverse datasets. Onehouse aims to abstract low-level data infrastructure so AI teams can focus on models while the platform handles ingestion, transformation and optimization. It is introducing two new products—Onehouse LakeView, a free observability tool for table stats and workload trends, and Table Optimizer, a managed service to speed ingestion and transformations. The startup emphasizes an open and interoperable approach to avoid vendor lock-in and make a single copy of data accessible across tools like Databricks, Snowflake and cloud-native services. With growing demand for lakehouse architectures to power real-time analytics and GenAI workloads, Onehouse positions itself to reduce storage and processing costs while improving query performance. Onehouse emerged last year with a cloud data lake product built on top of Apache Hudi. The company positions itself as a neutral integration layer to move data between different repositories and interoperate with multiple formats, rather than competing directly with vendors like Snowflake and Databricks. Its thesis is that many engines will sit on the same data and organizations should avoid migrating every time a new engine appears. The open source Hudi project has seen notable traction, with roughly 2,200 pull requests on GitHub last year and more than 600 engaged users; the company expects some open-source users to convert to its managed service. The cloud product remains in preview while Onehouse throttles new customers, though it already has paying customers in production. With recent funding the company doubled headcount from 15 to 30 and plans to focus hiring on go-to-market as it scales. Onehouse is building an open-source data lakehouse and a managed offering based on the Hudi project. Founder and CEO Vinoth Chandar created Hudi at Uber in 2016 and donated it to the Apache Software Foundation in 2017; he remained chair of the project and launched Onehouse early last year after a brief stint at Confluent. The startup raised an $8 million seed from Greylock and Addition to begin building its managed product and grow the business. Onehouse has 15 employees and plans to double headcount this year. The company intends to use the seed capital to expand the team, build out the managed service, and continue contributing to the Hudi open-source project. Hudi has drawn adoption from companies including Amazon, Walmart, GE Aviation, Robinhood and TikTok, which are building massive, exabyte-scale data lakes with the technology.

Total raised

$68M

Funding rounds

3

Latest round

Series B

Latest activity

Jun 2024