Engagement & pricing
Fixed, written quotes. No open-ended meters.
Every build is priced against an agreed specification after a short discovery. You know the number before we start, and AI running costs are billed to accounts you own.
-
Discovery Sprint
1–2 weeks
Quoted per scope
Map the workflow, test feasibility on your real data, and leave with a costed build plan.
- Process and data walkthrough
- Working proof-of-concept on a slice of real data
- Architecture, risk and cost notes
- Fixed quote for the build
- Most common
Fixed-Scope Build
3–10 weeks
Quoted per scope
A production system delivered against an agreed spec, with evaluation, docs and handover.
- Agreed spec and acceptance tests
- Weekly demos on a staging environment
- Evaluation suite and monitoring
- Deployment, runbook and handover session
-
Monthly Retainer
Rolling monthly
Quoted per scope
Ongoing improvements, new automations, model refreshes and on-call support for what we built.
- Reserved engineering capacity
- Prioritised backlog you control
- Monitoring and incident response
- Monthly report of changes and costs
-
Advisory & Audit
Hourly blocks or a fixed audit
Quoted per scope
Architecture reviews, AI readiness audits and second opinions for in-house teams.
- Written findings and recommendations
- Review calls with your engineers
- Vendor / model selection support
- Prioritised roadmap
-
Team Workshop
Half-day to 3 days
Quoted per scope
Hands-on training on your own use cases so your team can build and run AI safely.
- Pre-workshop needs survey
- Hands-on labs on your tools
- Templates and reference code
- Follow-up Q&A session
Want to understand typical effort first? Read our AI agent cost guide, RAG chatbot cost guide and fine-tuning cost guide.
FAQ
Pricing questions
Why don't you publish fixed prices for every service?
Because the cost of an AI project depends mostly on your systems, data and quality bar, not on the service name. A one- to two-week Discovery Sprint produces a fixed price against a written spec — the most honest number we can give.
Which currencies do you invoice in?
International clients are usually invoiced in USD; other currencies can be agreed. Payment terms are set out in the proposal.
How are AI running costs handled?
Model API, GPU and hosting costs are billed by the providers directly to accounts you own, so you see exactly what you pay. We estimate them up front and report them monthly.
Is there a minimum engagement?
The smallest engagement is an advisory block or a Discovery Sprint. Builds follow a sprint so that scope and price are grounded in evidence.
What payment schedule do you use?
Typically a deposit at kickoff and milestones tied to accepted deliverables for builds, and monthly in advance for retainers. Details are in each proposal.
Start a project
Have a workflow that should run itself?
Tell us what you want automated or built. You get a written assessment of feasibility, risks and a fixed quote — usually within two business days.