▪ The Six Phases
From ambition to governed production.
01
Discovery
2–4 weeks
Understand the problem before proposing the answer. We map the business problem, the deployment surface, the stakeholders whose sign-off matters, and the constraints. The output is a one-page engagement thesis both sides sign.
Stakeholder mapping and interview series
Problem framing and hypothesis set
Success criteria defined in business terms
Risk register first pass
02
Assessment
3–5 weeks
Measure readiness against the actual use cases, not in the abstract. Four questions per use case: is the data in usable form, do the platforms and MLOps exist, is the workforce capable, is governance sufficient.
Data readiness: lineage, freshness, access, quality
Technical readiness: platforms, MLOps, observability
Workforce readiness: literacy, fluency, coverage
Governance readiness: policy, risk tiering, owner model
03
Strategy
Sequenced
Prioritise, sequence, and commit. We score candidate initiatives on value, feasibility, time-to-value, and risk, then sequence a 12–24 month roadmap with named owners, dependencies, and decision gates. Written to be approved, not just presented.
Use-case prioritisation and portfolio design
Operating model and funding structure
12–24 month roadmap with named owners
Board-ready investment narrative
04
Build
8–16 weeks
Engineer the systems as production software. Build starts with the evaluation harness, not the model — real tasks, real data, graded rubrics — with guardrails, observability, and rollback discipline from the first commit.
Evaluation harness and graded task set
Tool and integration design with rollback
Guardrails and policy enforcement layer
Observability, logging, and incident triggers
05
Deploy
4–8 weeks
Shadow, supervised, autonomous — with explicit promotion criteria. Shadow runs alongside the current process with no customer impact. Supervised promotes it under human oversight. Autonomous is granted only when the harness, policy layer, and monitoring all pass.
Shadow → supervised → autonomous promotion gates
Runbooks, on-call, and incident playbooks
User and operator enablement
Monitoring dashboards signed off with the client
06
Govern
Ongoing
Operate the system under continuous governance. Production is not the end — it is when the clock starts. Alignment is to the frameworks that actually bind the organisation: EU AI Act, NIST AI RMF, ISO 42001, and NDPA 2023.
Drift and performance monitoring tied to KPIs
Retraining schedule and triggers
Incident response and post-mortem discipline
Audit evidence: model cards, DPIAs, attestations
▪ Engagement Shapes
Sized to the problem, not the invoice.
8–14 wks
AI Strategy
A board-approved roadmap written backwards from production — named use cases, owners, and investment envelopes.
10–16 wks
Production Agent
A well-scoped workflow taken from kickoff to supervised production, with an evaluation harness and governance built in.
12–18 mo
Transformation
Multi-year programmes run in tranches — capability, systems, and institutional knowledge that stay in-house.
We price per engagement against scope, duration, and the seniority mix required. Strategy engagements are fixed-fee; build and transformation run as fixed-fee phases with milestone-based release. We sign mutual NDAs at a client's request before any substantive discussion.
