How we deliver every engagement
Four steps, the same on a two-week assessment and a three-year managed service. Outcomes are agreed first, and we report against them to the end.


Step 1 of 4, 2–4 weeks
Discover
Two to four weeks with your business and technology teams. We map the estate, the data and the constraints, and agree measurable outcomes before scoping any build.
What happens
- Interviews with the people who run the process today
- Estate, data and integration mapping
- Baseline measurement of the numbers we will be judged on
- Security and compliance review
What you hold at the end
- Agreed outcome metrics and baseline
- Prioritised use cases
- Costed delivery roadmap

Step 2 of 4, 2–3 weeks
Design
Architects and domain leads set the target architecture, delivery plan and team shape, with security, compliance and run costs signed off up front.
What happens
- Target architecture and integration design
- Evaluation plan for any AI component
- Team shape, sprint plan and release path
- Run-cost model signed off with your finance and security leads
What you hold at the end
- Architecture decision records
- Delivery plan and team charter
- Security and compliance sign-off

Step 3 of 4, 3–9 months
Build
Onshore leads and offshore pods ship in two-week sprints against your real systems, with automated testing and, for AI work, evaluation gates before release.
What happens
- Two-week sprints with a working demo at the end of each
- Automated tests in the pipeline from the first commit
- Evaluation gates before any model reaches users
- Phased releases that follow your change calendar
What you hold at the end
- Working software in your environment
- Test and evaluation reports
- Runbooks written as we go

Step 4 of 4, Ongoing
Run & Improve
We hand over with runbooks, or run it for you under SLA. Either way we track the outcomes agreed in discovery and review cost, performance and models every quarter.
What happens
- L1 to L3 support under agreed service levels, or a structured handover
- Monitoring for availability, cost and model drift
- Monthly service reviews
- A quarterly improvement plan against the discovery metrics
What you hold at the end
- Service reports against the SLA
- Quarterly improvement plan
- Updated runbooks and architecture records
Ways to Work With Us
Discovery & Assessment
2–4weeks, fixed fee
A short, fixed-fee engagement to assess your estate, size the opportunity and leave you with a costed roadmap, whether or not we build it.
- Stakeholder & Process Workshops
- Architecture & Data Assessment
- AI Use-Case Prioritisation
- Security & Compliance Review
- Costed Delivery Roadmap
- Executive Readout
Quality and security are part of the build
These practices apply on every engagement, from a two-week assessment to a multi-year managed service.
- ISO/IEC 27001:2022
- Our information security management system is certified to ISO/IEC 27001:2022, and every engagement runs inside it.
- Your data stays with you
- Client data stays in your environment wherever possible. Access is role-based and logged, and we never use your data to train models for anyone else.
- Compliance as each engagement needs
- We work to GDPR, HIPAA or SOC 2 requirements as the engagement calls for, and agree the controls in the design step, not after the build.
- Evaluation gates for AI
- A success metric and an evaluation set are agreed in discovery. A model reaches production only when the metric holds, with monitoring and human review built in.
- Tests in the pipeline
- Automated unit, integration and regression tests run on every change, so quality is checked continuously rather than at the end.
- You own everything
- Code, models, prompts, documentation and data produced under your engagement are assigned to you.
See how this played out in practice in our digital bank case study.
Questions Worth Asking First

Still have a question?
Ask us about scope, engagement models, security reviews, or whether AI is even the right answer for the problem you have.
Talk to our teamFixed-fee discovery, fixed-price or time-and-materials projects, dedicated Offshore Development Centres, and SLA-backed managed services. Many clients start with one and move to another as the work matures.
Typically 15–20% onshore: engagement leads and architects who work close to your teams. The rest of the team works from our offshore delivery centres.
Teams run staggered shifts, so you get at least four hours of daily overlap whichever time zone you work in. Production support is 24x7.
You do. Code, models, prompts, documentation and data produced under your engagement are assigned to you, and we never use your data to train models for anyone else.
Our information security management system is certified to ISO/IEC 27001:2022. Client data stays in your environment wherever possible, access is role-based and logged, and we work to GDPR, HIPAA or SOC 2 requirements as each engagement needs.
We agree the success metric and an evaluation set during discovery, pilot against your real data, and promote to production only when the metric holds. MLOps, monitoring and human review are built in from the first release.
Usually with a call with one of our practice leads, followed by a two-to-four-week discovery. You get a costed proposal at the end of it, whether or not we do the build.
Ready for the first step?
A call with a practice lead, then a two-to-four-week discovery and a costed proposal.