Senior people, small teams and fixed scopes where the work allows it. Pick the model that matches how much you already know, then we write it down before anyone starts.
If the outcome is clear but the solution is not, start with discovery. If both are clear, fix the scope. If the roadmap keeps moving, bring in a team. If you need leadership more than hands, take a fractional CTO. If it is already live, let us run it.
For deciding what to build and whether it is worth building. Use-case selection, data readiness, architecture and an estimate.
Typically 1–3 weeks · fixed fee
AI strategy and consultingFor a system with clear outcomes and acceptance criteria, usually following discovery. Priced per milestone.
Typically 6–16 weeks · fixed price per milestone
AI implementationFor product companies and programmes where the backlog is yours and the capacity is ours. A named lead and two to six engineers.
Typically 3 months or more · monthly
Custom software developmentFor organisations that need senior AI judgement a day or two a week: strategy, vendor decisions, governance, hiring and architecture review.
Typically 1–2 days a week · monthly retainer
Fractional AI CTOFor live systems that need monitoring, upkeep and small improvements: AI applications, Salesforce orgs, event platforms and custom software.
Ongoing · monthly
Salesforce managed servicesThirty minutes with a practice lead is usually enough to say which model fits. If none does, we will say so and suggest who might be better placed.
Book a call| Criteria | Discovery sprint | Fixed-scope build | Dedicated team | Fractional AI CTO | Managed services |
|---|---|---|---|---|---|
| Best for | Deciding what to build | Clear scope, clear acceptance | Evolving roadmap | Senior judgement, part-time | Live systems |
| Typical duration | 1–3 weeks | 6–16 weeks | 3+ months | 3+ months | 12-month terms |
| Commercial model | Fixed fee | Fixed price per milestone | Monthly capacity | Monthly retainer | Monthly, with allowance |
| Team shape | Lead + architect | Lead + 2–5 engineers | Lead + 2–6 engineers | One senior person | Shared support team |
| Who owns the backlog | Shared | Fixed in the SOW | You | You, with our input | Shared |
| Change requests | Not applicable | Change control | Re-prioritise freely | Not applicable | Within allowance |
| Ending it | Ends with the deliverables | Ends at acceptance | Notice period | Notice period | Notice period |
| Fixed price available | Yes | Yes | No | No | Yes, per term |
What will be true when this is done, and how will we measure it? If we cannot answer, we are not ready to estimate.
Each with acceptance criteria you could hand to a tester. Vague deliverables are where estimates die.
The engineers who will do the work give optimistic, likely and pessimistic figures. Sales does not estimate.
Integration unknowns, data quality, model evaluation and your availability are listed as separate lines, not hidden in a margin.
A range with the assumptions that drive it, and usually a smaller version that proves the idea sooner.
We decline fixed prices for work that has not been through discovery, AI features with no evaluation plan, and timelines that depend on a third party we cannot influence. A smaller, honest scope beats a large, hopeful one. Our AI readiness assessment is a quick way to see whether you are at the discovery stage or the build stage.
Every engagement, however small, has a statement of work. It is short, written in plain English and designed to be read by the people doing the work as well as the people paying for it.
A decision-maker who can attend a demo every two weeks, access to the systems and data in scope within the first week, and a single place for questions. Most delays we have seen were waiting, not building.
Workshops, sprints, QA and support, as we have run them since 2006.
How we workSix engineering case studies from the products we run ourselves.
Case studiesYes, and most clients do. A discovery sprint usually leads to a fixed-scope build, which becomes a dedicated team or managed services once the system is live. Each change is a new statement of work, not a renegotiation of the old one.
Dedicated teams and fractional AI CTO work are billed monthly for agreed capacity. Builds are fixed price per milestone where the scope is clear enough; where it is not, we run a discovery sprint first rather than guess.
A discovery sprint of one to three weeks. It produces a written spec, an architecture and an estimate you can take to anyone, including another supplier.
You own the deliverables and the code we write for you, on payment. We retain our pre-existing tools and know-how, and we tell you up front if a proposal includes licensed components such as Eventify.
Small changes are absorbed within a sprint. Anything that moves a milestone goes through change control: we re-estimate, you decide, and the statement of work is amended in writing. No surprises on the invoice.
Tell us about your systems, your data and your constraints. We reply within one business day with a straight answer and a suggested next step.