OneHouse

OneHouse is a managed, universal data lakehouse platform built on Apache Hudi and supporting multiple open table formats (Hudi, Iceberg, Delta Lake).

Reviewed by 7wData

On this page

Publisher review

OneHouse is a managed, universal data lakehouse platform built on Apache Hudi and supporting multiple open table formats (Hudi, Iceberg, Delta Lake). Founded in 2021 by Vinoth Chandar—the original creator and PMC chair of Apache Hudi—the company provides fully managed infrastructure for data ingestion, transformation, and analytics. The platform operates as a SaaS offering deployed within customer VPCs on AWS, GCP, or Azure, ensuring data remains in customer-controlled storage buckets.

OneHouse addresses the operational burden of maintaining data lakehouses by automating table optimization, incremental processing, schema evolution, and data quality validation. Its OneFlow service handles change data capture from PostgreSQL, MySQL, SQL Server, and MongoDB, plus Kafka event streams and cloud storage files. Performance claims include 4x faster Spark execution via the Quanton compute runtime and 2-30x query acceleration through automated table optimization.

The platform includes LakeBase for SQL queries, LakeView for metadata exploration, table cost analysis, and support for multiple query engines (Trino, Ray, Presto). Integration with dbt is available for teams using dbt transformation workflows. Founded with $68 million in Series A and B funding from investors including Greylock Partners and Addition, the 68-person company in Menlo Park competes with Fivetran, Matillion, Astronomer, and Dataiku. OneHouse also launched Open Engines, allowing deployment of Apache Flink, Trino, and Ray on top of the lakehouse platform, targeting organizations seeking managed infrastructure without vendor lock-in to proprietary formats.

Get the AI & data signal, daily.

335k+ subscribers read this every morning. One email, both newsletters. Unsubscribe anytime.

How it works

  1. OneFlow Managed Data Ingestion

    Fully managed, serverless data ingestion with change data capture, Kafka integration, cloud file loading, and incremental processing to reduce compute costs.

  2. Multi-Format Table Support

    Native support for Apache Hudi, Iceberg, and Delta Lake with automatic metadata translation via Onetable to prevent vendor lock-in.

  3. Quanton Compute Runtime

    Proprietary Spark and SQL engine claiming 4x cost-performance improvement over standard Spark on customer infrastructure.

  4. Automatic Table Optimization

    Hands-off indexing, compaction, and partitioning that claims 2-30x query acceleration without manual tuning.

  5. Multi-Cloud VPC Deployment

    SaaS management layer deployed in customer VPCs on AWS, GCP, or Azure; data never leaves customer-controlled storage.

  6. Open Engines Framework

    Deploy Apache Flink, Trino, and Ray directly on OneHouse infrastructure with automatic connectivity to lakehouse tables.

Strengths and trade-offs

Strengths

  • Built and led by the original Apache Hudi creator, with deep expertise in lakehouse design and data format interoperability.
  • Consumption-based pricing (per compute-hour) aligns cost with usage; 1-month free trial with $1,000 credits mitigates setup risk.
  • Fully managed service eliminates operational burden of cluster provisioning, scaling, and maintenance; incremental processing claims 20-80% cost savings vs. full-scan ETL.

Trade-offs

  • Onboarding and documentation are reported as dense or overwhelming; learning curve for teams new to open table formats or the lakehouse model.
  • Pricing model lacks transparency on per-unit costs; contract-based structure makes budget forecasting difficult without vendor contact; additional AWS infrastructure costs not itemized.
  • Relatively small team (68 people) and no Gartner or Forrester positioning; market adoption and production case studies are limited compared to Fivetran or Matillion.

Pricing context

OneHouse uses consumption-based pricing: customers pay per Onehouse Consumption Unit (approx. $0.01 per unit based on AWS Marketplace listing), billed on compute-hours used to deliver ingestion, query, and optimization services. Pricing is contract-based (monthly or longer terms); upfront or installment options available. A 1-month free trial with $1,000 in credits is offered to approved customers.

Additional AWS/GCP/Azure infrastructure costs apply separately. Custom contracts and alternative commercial terms are available through sales contact (gtm@onehouse.ai). No published per-feature or per-seat pricing; all prospecting is sales-driven.

Alternatives

User reviews

No user reviews yet. Be the first to write one.

Sources

Reporting on this tool draws on these publicly available sources.

  1. www.onehouse.ai — Core features, pricing model (consumption-based), deployment (VPC), and cost savings claims (20-80%)
  2. www.onehouse.ai — OneFlow data ingestion capabilities, data sources (Kafka, CDC, cloud files), serverless autoscaling, pricing model
  3. tracxn.com — Founding date (2021), headquarters (Menlo Park), funding ($68M Series A/B), team size (68), founder (Vinoth Chandar), competitors (Fivetran, Matillion, Astronomer, Dataiku, Nexla)
  4. aws.amazon.com — Deployment model (SaaS), pricing structure (contract-based, pay-per-compute-unit), free trial terms (1 month, $1,000 credits), licensing (EULA-based)
  5. www.techbible.ai — Key features (ingestion, table optimization, multi-format support, LakeBase, LakeView), use cases, integrations (dbt, BigQuery, Snowflake, Databricks), alternatives
  6. aitools.xyz — User feedback on strengths (continuous data delivery, faster data preparation) and weaknesses (onboarding, feature overload)