FAQ

Everything people ask us.

Short answers, no hedging. If your question isn't here, ask us.

Deciding

Couldn't we just build this ourselves?
You could. By next year you'd have a platform team, an internal SDK, and a new pile to govern. Teams that start with a bare JupyterHub usually learn the same lesson: the notebooks were never the hard part. The catalog, access control, deployment, and audit trail around them are. Buying the floor is cheaper than building it.
What does it cost?
A flat annual license for the whole organization. The number depends on company size, not on usage. The pricing page has the details.
What is AdaLab, in one paragraph?
A governed home for everything your teams build with AI and data. It runs inside your infrastructure, catalogs every app and model, gives builders a browser-based workspace, and lets them ship to colleagues with access control attached. One flat license covers the whole organization.
Which industries already use it?
Pharmaceutical research, financial services, and research institutions, primarily. Sectors where reproducibility and auditability are requirements.
Are we locked in?
No. AdaLab is built on open-source components and open file formats. Your code is plain git, your data stays in your systems, and nothing proprietary stands between you and leaving.

Running it

Does our data leave our infrastructure?
No. AdaLab deploys into your cloud or on-prem environment. Prompts, data, and logs stay inside your network.
How is it deployed?
Kubernetes-based, in your cloud (AWS, Azure, GCP) or on-prem. Your team keeps control of the perimeter; we help with the rollout.
Do our builders have to change how they work?
No. They keep their editors and notebooks. AdaLab is where the result lands, not a new way to write code.
Can business users ship apps without IT?
Yes. Anyone can deploy an app or dashboard from their workspace, inside the governance boundaries IT has set. IT defines the rails once; people ship without tickets.
How does it connect to our existing data systems?
Pre-built connectors, SQL support, and custom integrations, with role-based access per source. Each connection is made once and reused everywhere it's allowed.
Does it support FAIR data principles?
Yes. Central access controls, consistent metadata, open file formats, and full data lineage. Built for sectors where FAIR is a requirement, not a preference.

If your question isn't here, ask us. Ask us directly