The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Onehouse announced a $35 million Series B on June 26, 2024, led by Craft Ventures, with existing investors Addition and Greylock Partners participating. The company said the round brought its total funding to $68 million and would support its managed lakehouse platform, open-source work and expansion of its engineering and go-to-market teams. The announcement also introduced LakeView and Table Optimizer. This is a 2024 funding event, not a new financing announcement.
Why the funding matters
Onehouse is trying to address a persistent data-platform trade-off: warehouses offer managed operations and familiar analytics, while data lakes can store large volumes of data in cloud object storage and support a range of processing engines. A lakehouse aims to combine those approaches, adding table management and transactional features to data stored in a lake. The term describes a family of architectures, not one standardized product category.
The company’s pitch is to keep data in open table formats and customer-controlled cloud storage while managing much of the work needed to ingest, transform, optimize and query it. That approach may appeal to teams seeking more engine choice without building every part of a lakehouse themselves. It does not, by itself, establish that Onehouse is cheaper or more portable than a warehouse or competing managed platform.
What Onehouse said the money would fund
Onehouse said it would use the financing to develop its managed data lakehouse, advance interoperability and performance work, support Apache Hudi and Apache XTable, grow its development organization and expand go-to-market efforts. The company did not disclose a line-item allocation of the $35 million. Its founder and CEO, Vinoth Chandar, created Apache Hudi before founding Onehouse in 2021. Onehouse’s Series B post describes the company’s strategy and investment priorities.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problems#1 Best Overall
The announcement framed the funding as support for a commercial platform built on open-source foundations—not as funding for an open-source project alone. The company’s announcement named Craft Ventures as lead investor and Addition and Greylock Partners as participants.
What “open” means—and what it does not
Onehouse emphasizes open table formats, multiple catalogs and query engines, and storage that remains in a customer’s cloud environment. Its current site says the platform supports Apache Hudi, Apache Iceberg and Delta Lake, can run in AWS, Google Cloud and Azure, and can be deployed in a customer’s own VPC. These are Onehouse’s product claims, not independent verification of every configuration or workload. The current Onehouse site describes its positioning.
Open formats can make it easier to give different engines access to shared data and may reduce dependence on a proprietary storage layer. They do not guarantee a frictionless exit from a managed service. Orchestration, security configuration, catalogs, monitoring, proprietary optimizations, support relationships, staff expertise and cloud egress charges can all create practical switching costs.
Rank #2
Hudi, Iceberg and Delta Lake
Apache Hudi is central to Onehouse’s history and the incremental data-management problem its founder worked on at Uber. Apache Iceberg and Delta Lake are other widely used table formats in lakehouse architectures. The formats overlap, but their capabilities and implementation details differ. Onehouse positioned Apache XTable as an interoperability layer for translating or synchronizing table metadata across formats; that should not be read as a guarantee that every feature or table can be converted losslessly.
Free tools Windows power users keep installed
One-click scans. No signup required.
- Transaction behavior and how concurrent changes are handled can differ.
- Deletes, updates, schema evolution, partitioning and time-travel features may not map identically.
- A query engine must support the target format and the particular features a workload uses.
- Ask whether a format transition changes metadata only or rewrites data, and test the behavior on representative tables.
The products announced with the Series B
LakeView: table observability
At launch, Onehouse described LakeView as a free observability service for lakehouse tables. The announcement listed table statistics and trends, timeline history, partition-skew visibility, file-size distribution, compaction monitoring and alerts about inefficiencies. The launch description does not establish current feature limits, support terms or whether every feature remains free. Nor does it show adoption or revenue from the product. LakeView’s product page is the company’s current reference.
Table Optimizer: managed maintenance
Onehouse described Table Optimizer as a managed service for tasks including incremental clustering, asynchronous compaction of small files, cleaning data beyond time-travel retention periods and improving ingestion and query performance. The company claimed up to 10x better query performance and said one Snowflake/Iceberg use case could deliver up to 2x faster queries than ingesting data into external Iceberg tables for Snowflake. These are vendor-reported maximums, not independently established results for typical workloads; the announcement does not provide a general benchmark basis that readers can apply to their own systems. The Table Optimizer page describes the service.
Maintenance addresses a real operating problem. Small files, skewed partitions, frequent updates, excess metadata and retained historical data can increase storage, scanning and query overhead. Compaction and clustering can help, but use compute and may compete with ingestion or queries. The right schedule depends on read and write patterns, freshness targets, retention rules and cloud prices; optimization is a workload trade-off, not a free performance gain.
Onehouse’s managed-platform model
Onehouse is more than a table format. Its documentation describes managed clusters, OneFlow ingestion, Spark jobs, SQL pipelines, table optimization, catalog synchronization and access to open engines such as Trino, Flink and Ray, alongside its Lakegres query layer. The exact capabilities and deployment requirements should be checked against the current documentation rather than assumed from the 2024 announcement. Onehouse documentation provides the current technical reference.
The commercial proposition is managed operations around data that the company says can remain in a customer’s cloud environment. Buyers should confirm which cloud regions and services are supported, how identity and permissions integrate, who controls encryption keys, what private connectivity is available, and how backups, disaster recovery, service levels, egress and migration work. Public materials cited here do not establish every enterprise requirement or a public rate card.
Rank #4
How it compares with other approaches
| Approach | Potential advantage | Main trade-off |
|---|---|---|
| Onehouse managed lakehouse | Managed operations with an emphasis on open formats, multiple engines and customer-cloud deployment, according to Onehouse. | Buyers still need to assess service dependency, feature portability, pricing and operational fit in their own environment. |
| Warehouse-centric platform | A more integrated experience can suit teams focused on SQL analytics, established governance and fewer components to operate. | It may be less attractive to teams prioritizing broad multi-engine portability or direct control of the data architecture. |
| Integrated lakehouse platform | A broad platform can bring data engineering, analytics and machine-learning tools together. | Teams seeking a neutral layer across competing engines may see an integrated platform as another source of platform dependence. |
| Cloud-native services assembled by the customer | Composable services can offer granular control and close integration with a chosen cloud. | The customer takes on more integration, tuning and operational responsibility. |
| Self-managed open-source stack | Maximum architectural control and direct responsibility for the chosen components. | Teams must operate ingestion, orchestration, table maintenance, catalogs, security, observability, upgrades and incident response. |
These are different operating models, not a single feature checklist. Snowflake, Databricks and AWS offer distinct platform approaches; the relevant choice depends on existing skills, workloads, governance and how much operational work a team wants to retain. Onehouse’s position is most relevant when a company wants open-format data but also wants help running the lakehouse stack.
What buyers should test before choosing
- Portability: Test the specific formats, engines, catalogs and table features in use, including updates, deletes, time travel and schema changes.
- Total cost: Model managed-service charges alongside cloud compute, object storage, compaction, rewrites, network traffic and egress. No public list price for Table Optimizer or a complete platform rate card is established in the cited materials.
- Performance evidence: Request benchmarks using representative data sizes, file layouts, queries, concurrency and freshness targets; compare against a clearly specified baseline.
- Deployment and governance: Verify region availability, VPC and networking requirements, access controls, data residency, encryption and key management, support and SLA terms.
- Exit plan: Establish how to export metadata and data, replace managed pipelines or optimizations, and estimate the time and cost to move.
Where AI fits into the story
Onehouse and its investors positioned lakehouse infrastructure as useful for real-time analytics, predictive machine learning and generative-AI workloads. The company’s current website also describes vector-embedding generation and delivery to AI/ML platforms and vector databases. Flexible, reusable data foundations can matter as organizations add these workloads, but the financing announcement does not prove a distinct AI advantage for Onehouse.
What the $35 million announcement establishes
The Series B shows investor backing for Onehouse’s attempt to commercialize a managed service around open-format lakehouse data and interoperability. It does not prove that format translation is seamless, that performance claims generalize, or that the platform wins on price. The decision for a data team is whether it can retain enough control and portability while outsourcing enough operational work to justify another managed platform.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
