Architecture · Engagement 01
AI Discovery & Scoping
Decide what to build — and whether to build it at all.
Most failed AI programmes were mis-scoped before a line of code was written. Discovery replaces the enthusiasm with a defined problem, tested feasibility, costed options and a delivery approach you can put in front of a board.
- Problem definition and the decision the system is meant to support
- Success criteria, and what would count as evidence of them
- Data availability, quality, rights and access constraints
- Technical feasibility assessment against the stated criteria
- Build, buy and hybrid options with their trade-offs made explicit
- Indicative cost envelope and run-rate drivers
- Delivery approach, sequencing and the capability required to execute it
- Early risk, governance and regulatory flags
- Discovery report with the problem statement and success criteria
- Feasibility assessment and evidence base
- Costed options analysis with a recommendation
- Delivery roadmap and resourcing shape
- Risk and governance flag register
- Executive read-out session