Project delivery
A delivery method built around clear review points
Move from discovery to support with clear records, owners, checks, and sign-off.
How projects run
Define the review points before work begins
The Data-Driven AI Project Methodology runs from discovery and design, through build and test, to handover and support. Each phase has a stated goal, a working record, and a review before the next one starts.
Project steps
The lifecycle keeps decisions visible
Each project sets its own outputs. These records keep the scope, risks, design, tests, and owners clear as work moves on.
Phase 01
Discovery and assessment
Learn how things work now. Meet the key people. Agree the problems, KPIs, users, data needs, and signs of success.
Working record: project charter, scope statement, RAID register, and first architecture review.
Phase 02
Solution design
Plan how the team will bring in, store, use, govern, report, and share data. Check the plan against the security, compliance, and architecture needs in scope.
Working record: high-level and low-level designs, architecture diagrams, and the security and compliance plan.
Phase 03
Implementation and testing
Build and test the agreed work in short sprints. Use plans, reviews, test proof, and feedback to check it against the acceptance criteria.
Working record: configured environments, tested pipelines or reports, and sprint reports.
Phase 04
Handover and documentation
Give the team the guides, training, files, and review proof set in the scope. Share the knowledge they need to run the work.
Working record: handover files, user guides, and the review after launch.
Phase 05
Maintenance and support
Agree the support after launch. Set service levels, review dates, ways to raise issues, and the move into day-to-day work.
Working record: support reports, health checks, and a plan for later improvements if they are in scope.
Project checks
Governance lives in the working record
Keep roles, risks, sign-offs, and test proof clear from start to finish.
Team rhythm
Turn each request into a decision record
The team may include an account manager and data architect. It may also include engineers, data model and governance experts, report analysts, and change advisers. The scope sets the final team, meeting plan, and support.
- Document the request, intended outcome, scope, constraints, and timing.
- Agree the proposal, deliverables, roles, communication cadence, and escalation path at kickoff.
- Record progress, risks, dependencies, decisions, and review evidence in the agreed tools.
- Complete acceptance, documentation, and knowledge transfer before agreeing any ongoing support terms.
Put review points around the project
Bring your current setup, goal, limits, and the proof each owner needs.
Want to learn more?
Explore the rest of the site or get in touch with the team.