The Bespoke Build Process
Discovery, Build, Handoff, Run. Four stages from your first call to a working system your team owns, each with a timeline and a clear deliverable, so you always know what happens next and what it costs.
Why the process has a name.
Most AI projects do not fail in the build. They fail at handoff, when the team that has to live with the system never learned how it works, or they drift after launch, when an API or a model changes and nobody is watching. The Bespoke Build Process exists to close both gaps: a fixed-price scope before anything is built, a handoff that ends with your team able to run the system, and an optional Run tier for businesses that would rather we kept it working.
The same four stages apply to every project, whether it is a single workflow, agent or internal tool, and the price is set by scope: projects start at $2,000.
Discovery
We start with a 30-minute automation review. You walk us through your business, the tasks that slow your team down, and what you would like to see automated. We ask the questions you may not have considered: where the information lives, who touches it, what breaks when someone is away, and which steps genuinely need a human decision.
Within 3 business days you receive a written scope: exactly what we will build, how it works, how long it takes and a fixed price. We walk through it together, you push back on anything that does not feel right, and nothing starts until we both agree. If you want to phase the build or simplify it, that happens here.
Most agencies scope exactly what you asked for. We scope what you asked for, and flag everything else we identified that could make your business run better. You decide what to act on.
What you receive
- Scope of work: clear definitions of what is being built
- Timeline: an exact schedule for milestones and delivery
- Investment: a fixed price with no vague estimates
- Additional opportunities identified, for you to act on now or later
Build
We build on the tools you already use. No migration, no new platform for your team to learn. A project manager keeps an internal timeline and keeps you updated throughout, with check-ins sized to the project: more frequent for larger builds, lighter-touch for smaller ones.
You do not need to manage the build. You need to be available for the check-ins, give us access to the tools the system touches, and tell us when something is not quite right. Projects over $10,000 are billed in milestones agreed in the scope, so you never pay for work you have not seen.
What you receive
- A named project manager on your engagement
- Progress calls at a cadence set at kickoff
- A working milestone you can test at each check-in
- Clear communication, with no unexplained delays
Handoff
When the build is complete, we do not just send you a file. We walk you through exactly how the system works, how to use it day to day, and what to do if something behaves unexpectedly. You receive full written documentation, so anyone on your team can understand the system without needing to call us first.
"Done" means the system runs on your accounts, against your real cases, with your team able to operate it. For 30 days after handoff we fix anything that does not match the scope at no charge.
What you receive
- Working system, fully tested against your real cases
- Live walkthrough session, recorded if you want it
- Written documentation your team owns
- 30-day support window
Run
Most AI projects do not fail at launch. They drift: an API changes, a model is updated, a prompt that worked in March misfires in June. Run is the optional monthly management tier that catches that before your team does.
We monitor every agent and workflow we built for you, fix them when a connected tool or model changes, tune prompts and rules from reviewing real outputs, handle a monthly allowance of small changes, and send a written monthly report of what ran, what we changed and what we recommend next. Hosting, SaaS licences and API usage stay in your own accounts and are not included.
What you receive
- Monitoring and alerting on every system we built
- Fixes and retests when an API, model or tool changes
- Prompt and rule tuning, plus a monthly allowance of small changes
- A monthly written report
Four things people ask before they start.
Not answered here? Book the free 30-minute review →
- Can we phase the build?
- Yes. Larger scopes are split into milestones you approve one at a time, so you can start with the piece that saves the most hours and add the rest once it has paid for itself. Each milestone has its own fixed price.
- What do you need from us during the build?
- Access to the tools the system touches, one person who can answer "how does this work today" questions within a day or two, and attendance at the milestone check-ins. We handle the rest.
- What happens if something breaks after handoff?
- Every build ships with error handling and alerts, and the documentation covers the common fixes. For 30 days after handoff we fix anything that does not match the scope at no charge. After that, the Run tier covers monitoring and fixes; without it, you can call us for a one-off fix at a fixed price.
- Is Run required?
- No. Every handoff includes documentation and a walkthrough so your team can run the system. Run is for businesses that would rather we did, at $299 per month, month to month. More about Run →
Need us to keep it running?
Some clients manage their systems internally after handoff. Others want the peace of mind of ongoing monitoring and maintenance. Run is our optional monthly management tier: a flat $299 per month, billed monthly in advance, cancel with 30 days' notice.
- Monitoring and alerting on every system we built for you
- Fixes when an API, model or connected tool changes
- Prompt and rule tuning, and a monthly allowance of small changes
- A monthly written report