A four-phase journey from your first conversation to your model in production.
We do not sell AI licences and leave you to figure the rest out. Every engagement starts with a structured assessment, moves to a measured pilot, and ends with a model your institution owns — governed, benchmarked, and running in your environment.
Start with an assessmentYou own. We license. We do.
Four phases. Clear milestones. Your model at the end.
Readiness assessment
Workshops with your business, risk, and technology leads. We review your data, map your regulatory gaps, and agree your top three use cases.
Who joins from your side
Business head for the first use case · Chief Risk or Compliance Officer · CTO or CIO
What you get
Use-case priorities, data-readiness view, regulatory gap map, pilot plan.
Pilot
One use case goes live on Vaman base models inside your environment. We measure against the scorecard agreed in your assessment.
Who joins from your side
Use-case owner · IT team · Compliance representative
What you get
Live agents on one use case, benchmark scorecard, go/no-go decision for production.
Production
Your model is trained on your production data, benchmarked against your agreed threshold, and deployed in your cloud or on-premise environment.
Who joins from your side
IT and security · Risk · Operations team
What you get
Your model, governed and running. Every action on the record.
Scale
Additional use cases added on the same platform. No new governance build required — your existing Guard rules and Ledger audit trail extend to each new agent.
Who joins from your side
Business heads for each new use case
What you get
More use cases, same platform, same audit trail.
Clear lines. No surprises.
Your readiness assessment ends with a priced proposal. We do not publish standard prices because every engagement depends on use-case scope, data readiness, and deployment environment.
Model build
A one-time fee per model tier (Ground, Tune, or Forge), covering training, validation, and benchmark review.
Platform licence
An annual fee for the Vaman Stack, by use cases and deployment environments.
Cloud compute
You pay your cloud provider directly. We never mark up compute costs.
Services
Fixed-scope projects or a managed service, plus annual maintenance and support.
Your IP, in plain English.
The model trained on your data is yours, including the weights.
Your data is used only for your models — never for another client.
The Vaman platform remains Vaman's and is licensed to you.
If you leave, you keep your models and data. The platform is covered by code escrow.
Questions about working with us
Is this SaaS?
Partly. The Vaman platform is licensed like software — you pay an annual fee and receive updates every quarter. But your model is yours, runs in your environment, and you keep it if you leave.
Is this a services company?
Partly. We build and run with you. But every client runs on the same platform, so you benefit from product improvements — not one-off code written for your institution.
Can our team run it ourselves over time?
Yes. You can start with our managed service and take over more operations as your team grows. The assessment will include a question about your internal AI ambition.
What if we do not go ahead after the assessment?
You keep every output from the assessment — use-case priorities, data-readiness view, regulatory gap map, and pilot plan. There is no obligation to proceed.
Find out what you'd own in three weeks.
The readiness assessment gives you use-case priorities, a data-readiness view, a regulatory gap map, and a pilot plan — whether or not you proceed.
Book A Readiness Assessment