About Hyperlake.
Hyperlake is the AI cloud for your team and your agents, on your cloud. It runs managed open-source stacks, from Kubernetes, ClickHouse and Postgres to Trino, Iceberg and open models, in the customer's own AWS or Google Cloud account, with LakeGate securing every query to its databases and data engines.
What Hyperlake does
The problem it solves
AI agents are becoming the main users of data and models. Teams that want open-source engines usually have two choices: run them themselves, or pay for the commercial edition, usually in the vendor’s cloud. Agents make both harder, because they multiply the queries, model calls and jobs, and with them the operating work and the usage bills. Hyperlake is the third option. We run the open-source stacks for you, inside your own cloud account.
How it is built
Open-source engines, installed in one run on a Kubernetes foundation, with LakeGate putting sign-in, roles and policy on every database query. There’s no markup on compute. Customer-cloud workloads run in your account, and the cluster can run with no public endpoint. Control-plane data and stored credentials may be processed elsewhere, as described in section 5.8 of the Terms.
Who it is for
Enterprise and platform teams
Teams that want to own their AI environment: their infrastructure, data, and governance.
AI product companies and service providers
Teams that deploy into many client environments and need repeatable installs, client ownership, and one operating approach.
Where it runs
Your infrastructure or your clients’
On AWS and Google Cloud today, with Azure, OVH, private cloud and on premises coming soon. The foundation runs on Kubernetes, which keeps services portable and manageable from one environment to the next. Governance and operations stay the same wherever the stack lives.
Talk to the team
Get in touch
Tell the Hyperlake team what you want to deploy, whose environment it must run in, and which data and policies it needs. Use the contact form, or read the FAQ and the use cases first. Existing users can sign in.
Start with a workload. Build the environment around it.
What do you need to deploy? Whose environment does it have to run in, and what does it connect to? Tell us.