Business ConsultingHome / AI Transformation / 10·20·30 Pathway
AI transformation roadmapDefine. Prove. Transform.
Our engineers work beside yours at every step. Each stage ends in evidence, not slides.
Three stages
Ten hours. Twenty days. Thirty weeks.
Discovery Lab
One focused session with your business, IT and data owners. We leave with success defined.
Who joins
- Business sponsor and process owner
- IT and data owners
- Risk, compliance or legal where decisions carry risk
- Vera consultant, data scientist and forward-deployed engineer
What we do
- Map the workflow and its decision points
- Write the enterprise problem statement
- Inventory data sources and agree a collection plan
- Set acceptance criteria and baseline KPIs
- Build the ROI model
What you get
- Signed-off problem charter
- Data collection plan and access list
- Acceptance criteria for the Proof of Value
- ROI model with baseline and target
Proof of Value
A working AI teammate. Your real data. Tested against the Lab's criteria. Shown to your stakeholders.
Week by week
- Days 1–5: data ingested, enriched and labelled
- Days 6–12: models trained and compared; rules derived from de-identified data
- Days 13–17: teammate assembled with human gates and audit trail
- Days 18–20: KPI test against the baseline and live demo
How it runs
- On IntelliMesh cloud, your private cloud or on-premise
- Your data stays in your control
- Forward-deployed engineers beside your team
What you get
- Working teammate on your data
- KPI results against acceptance criteria
- Validated ROI and scale-up plan
- Proposal for the 30-week programme
AI Transformation Programme
A signed programme. The proven teammate goes live, then scales into swarms across the business.
Phases
- Weeks 1–8: production hardening, integration and security review
- Weeks 9–20: roll-out across business units and sister crews
- Weeks 21–30: swarms at scale, managed run and handover
Included
- Production deployment: cloud, private cloud or on-premise
- Teammate swarms on one control plane
- Change management and team enablement
- Governance: RBAC, SSO, audit, approval gates
What you get
- AI teammates in production with measured KPIs
- Your models, rules and data, exportable
- Trained teams and an AI operating model
- Roadmap for the next wave of use cases
360 AIDLC
One lifecycle. Discovery to run.
The framework behind every stage of the Pathway.
Discover
Use-case and ROI assessmentData
Ingest, enrich and labelBuild
Train, compare and fine-tuneDeploy
Cloud, private or on-premOperate
Teammates run the workflowImprove
Feedback and retrainingTypical build 3–6 months → With IntelliMesh 2–4 weeks per use case · 27–30% lower cost in the first phase
Three routes, one journey
Which route is right for you?
Three ways in. One way through: a 20-day Proof of Value, then the 30-week programme.
| Route | Start here if | First step | What you get | Time |
|---|---|---|---|---|
| 10·20·30 Pathway | You know the process and want AI proven fast | 10-hour Discovery Lab | A working teammate on your data, then production at scale | 10 hrs · 20 days · 30 wks |
| Simplification | The process itself is slow, manual or broken | Lean Workshop & Dipstick, 3–4 weeks | Waste priced, the process redesigned, then the 20-day Proof of Value | ~11–14 wks to a proven AI use case |
| Business Advisory | You run a GCC and need a mandate, operating model and value story | GCC maturity diagnostic, 2–3 weeks | A board-ready roadmap, then a transformation office tracking benefits | 6–8 wks to the roadmap |
How engagements scale
Start fixed. Scale on outcomes.
30-week AI Transformation Programme
Production, teammate swarms and enablement across business units.
Process waves
Simplification at scale: a fixed fee per wave of 6–8 weeks.
Transformation office
Retainer for PMO, change and value tracking with finance.
AI teammates
Subscription, usage-based, enterprise licence or outcome-based.
Gain-share
Where savings are shared, our fees can be tied to them.
One benefits dashboard
Every route reports against the same baseline and KPIs.
Questions
Questions, answered.
How do we start?
Book a Discovery Lab through the contact form. It is a short, fixed-scope engagement; we will share current terms when we schedule it.
What data do you need for the Proof of Value?
Whatever the problem statement needs: documents, images, logs or system records. The Lab produces a data plan. Personal data is de-identified first. Everything stays in your control: our cloud, your cloud or on-premise.
What happens if the Proof of Value does not meet the acceptance criteria?
You will know within 20 days, with evidence. We agree the criteria up front. Pass or fail is never in doubt.
Who owns the models and rules?
You do. Models export as ONNX, PMML or H5. The rules from your data are explicit. And yours.
Start the 10·20·30 Pathway