An automation project with us runs in five steps: a plan and a fixed quote on day 1, access set up with the smallest permissions in week 1, build and test in weeks 2 to 4, launch with monitoring and a handover doc, then 30 days of fixes at no charge. Payment is split by project size and the written scope is agreed before anything is paid. Most single workflows go live in 2 to 6 weeks.
This post is the long version, so you know what to expect before you write to us. Every price and term here matches our pricing page as of October 29, 2026.
Before the project: the free audit
Most projects start with the free automation audit. You describe one manual task by email. We reply within one business day with one specific automation idea for it. There is no call required and no price in the audit reply. If the idea fits, we move to a plan.
If you want more than one idea mapped first, the AI readiness assessment is $3,199 over two weeks. We map your manual work, rank what to automate and give you a fixed quote. The full fee is credited to the build.
Step 1, day 1: the plan and a fixed quote
We go through the process with you on a call or in writing: what starts it, which tools it touches, what the normal case looks like and, most of all, the exceptions. The exceptions are where projects go wrong, so we ask for real examples of the awkward ones.
You get a written scope and a fixed-price quote. The price covers what the scope says. If you later want something the scope does not cover, we quote that separately before doing it.
Step 2, week 1: access with the smallest permissions
We ask for access to each system the automation needs, at the lowest level that works. Where a tool supports it, the automation gets its own user or API key, so its actions show up separately in your logs. Read-only access where reading is enough. We agree which test data to use before touching anything live.
If the work involves patient data, we list every vendor that would touch it so you can check each one has a business associate agreement with you, and where possible the automation passes record IDs instead of names or notes. Our security page explains how we handle access and data.
Step 3, weeks 2 to 4: build and test
We build against real examples from your systems, including the exceptions you gave us in step 1. For AI steps, we test on past cases and check the outputs by hand before anything goes live. Our post on testing AI agents before production shows the method.
Depending on the payment band, you see a working demo partway through, so you can correct course before the build is finished.
Describe one manual task. We'll reply within one business day with one idea to build.
Step 4, launch week: monitoring, alerts and a handover doc
We switch it on with monitoring in place. If a step fails, an alert goes to a named person, so a broken run does not sit unnoticed. Our post on automations that silently stop working explains why this matters.
You get a short handover doc: what the automation does, where it runs, how to pause it, and who to call. You own the build and the accounts it runs in.
Step 5, the 30 days after launch
If something we built breaks within 30 days of launch, we fix it at no charge. This is also when we tune things as your team gets used to it.
After 30 days: your choice
- Run it yourself. You own it. If you need a fix later, it is $250 an hour.
- Care plan, $350 a month. Uptime checks, fixes when a connected tool changes, and a monthly run report. Month to month with 30 days' notice.
- Managed agents. For businesses running several AI agents, managed plans start at $4,000 a month and cover hosting, monitoring and fixes. See AI agent services.
AI model usage is separate. For most small businesses it is about $30 to $150 a month, paid straight to the model provider with no markup from us.
How payment works
| Project size | Split | When |
|---|---|---|
| Under $1,000 | 100% upfront | When the work is booked |
| $1,000 to under $3,000 | 50% / 50% | Start and handover |
| $3,000 to under $5,000 | 25% / 50% / 25% | Start, first working demo, handover |
| $5,000 and above, and all client-side work | 33% / 33% / 34% | Start, 30% done, 60% done and handover |
An amount exactly on a boundary uses the higher band, so a $3,000 project is paid 25/50/25. Local clients pay 50% to start and 50% at handover. Payments are not refundable and there is no cancellation window, which is why the written scope is agreed before any payment.
What a project costs
Systems integration starts at $2,000, a single workflow at $2,500, custom development at $3,000, an MCP server at $3,500 and a first AI agent at $4,000. The cost guide shows what moves a quote up or down, with sample ranges such as $2,500 to $4,000 for a dental reminder system. If your business is a dental office or a real estate team, our dental package and real estate package have fixed prices and fixed scopes.
What we need from you
- One person who knows the process and can answer questions within a day or two.
- Real examples, including the awkward ones.
- Access to the systems, set up by whoever administers them.
- Someone to test it with us in launch week.
Projects slow down mostly when one of these is missing, so we raise it early if we see it.
Examples from client work
In our dental no-show case study, a reminder system built on n8n, Twilio and Dentrix went live in 3 weeks and cut no-shows from 18% to 7.4%. In our lead response case study, a real estate team's reply time dropped under 90 seconds, around the clock. Our benchmarks page collects the before-and-after numbers from all our case studies.
How this fits your next step
Pick the service that matches the work: workflow automation for one process end to end, systems integration to keep tools in sync, or AI agents when the task needs judgment. If you are not sure, the free automation audit is the place to start.
