Our work · Deployment
Frugal AI Deployment Labs
Deployment Labs work with public institutions and their partners on live AI challenges. The aim is to learn what it takes for a system to be deployed, governed and sustained in a specific place, and to leave that knowledge behind for the next adopter.
Why deployment
Adoption is a decision. Absorption is a capacity.
Adopting AI means deciding to use it. Absorbing it means the institution can fund it over its whole life, fit it into the work people actually do, govern it, and keep it running after the pilot team has gone.
Pilots tend to test the first. The Labs are built around the second.
How a Lab is set up
Four roles, clearly divided
Institutional owner
A government department or public body that owns the problem and the outcome, and stays accountable after the Lab.
Host site
The place where the work happens and where people will use the system day to day.
Technical partners
The organisations that build, integrate and operate the system.
The Hub
The methodology and measurement partner: problem framing, the sufficiency test, whole-life cost, evidence and governance. Building and operating the system stay with the technical partners.
How a Lab works
Six steps
- Frame the problemStart from the outcome, the people it serves and the constraints they work under.
- Test sufficiencyAsk what is enough for the outcome, including whether the workflow needs AI at all.
- Design for the conditionsConnectivity, language, budget, skills and governance shape the architecture.
- Measure from day oneCapture baselines at launch so cost, use and outcomes can be compared later.
- Decide what scalesScale what works. Stop what cannot justify its resources.
- Document the pathwayRecord the decisions so the next adopter starts further ahead.
Focus and status
Where the Labs work
Labs focus where public value is high and where tight budgets, patchy connectivity or many languages can make design assumptions visible early: health, agriculture, education, public services, climate and sustainability, and language and voice.
The first Lab is being set up in Andhra Pradesh, following a memorandum of understanding between Cambridge Judge Business School and the Andhra Pradesh Economic Development Board in November 2025. We will publish what we learn, including what did not work.
Propose a Lab
If your institution has a live AI challenge and a team that will own the outcome, we would like to hear about it.
