Services
AI automation consulting for the systems you already run
The same people scope the work and build it: an AI automation consultant who writes the agents, then runs them. Four services, one team, and our own company as the first customer for all of it.
What we do
Four services. Most engagements use two.
AI consulting for small business
We watch a week of the work, rank the jobs an agent should take, and quote the first one. No roadmap deck.
Custom AI agent development
One agent, one job, built on your systems and run by us. A documented process, a fixed price, and the code is yours at the end.
Custom AI solutions
Custom software and agents built around your requirements, hosted for your team, with a conversation to request the next improvement.
AI integration
Inject agents into the software you already use: Salesforce, QuickBooks, BILL, Slack and the rest, through their real APIs.
How it runs
What a first engagement looks like, week by week
Process automation consulting goes wrong in a predictable way: a long discovery, a deck, and nothing running. So the shape is fixed and short. You should know by the end of the first month whether this works, and you should have something in production rather than a roadmap.
- Week 1
Watch the work
We sit with the person who actually does the job, at the time they do it. Not a workshop and not a survey — the work itself, including the exceptions that never make it into a process document.
- Week 2
Map and rank
Every manual job written down by person and by hour, then ranked by time eaten against how safely an agent could take it. You get the list whether or not you build anything.
- Weeks 3–4
Scope and build one
One job, scoped in writing: what the agent will do, what it will never do, which systems it touches, what permissions it needs, what done looks like. Then it gets built against your accounts.
- From week 5
Run it
It goes live with an approval step and a kill switch. We host it, monitor it and read the logs, and you get a monthly report in plain English.
Scope
What we do not do
The list matters because most of what gets sold as AI automation consulting is one of those five things wearing a different name. An engagement that starts with a migration is a migration project, and the AI is the part that gets cut when it runs long. Ours starts with one job that a person is doing today, and the measure of it is whether that person stops doing it.
The approach
What a workflow automation consultant should leave behind
Start with the hours, not the tools
Process automation consulting usually starts from a product catalog. We start from a week of watching the work: who retypes what, who chases whom, what the phone does at 5:01 p.m.
Build inside the systems you already pay for
Your QuickBooks, Salesforce, Google Workspace and Slack stay exactly where they are. We connect through their APIs and put an agent in the gap between them.
Keep a person in charge
Every agent drafts first and acts second. Financial and customer-facing actions wait for an approval click until you decide otherwise. Logs for everything.
Questions we get asked
What does an AI automation consultant do that a software vendor doesn't?
A vendor sells you a tool and leaves the integration to you. An AI automation consultant looks at the whole workflow, picks the job worth automating, and is accountable for it actually running in your systems.
Is this workflow automation consulting or AI consulting?
Both, and the line has mostly disappeared. Workflow automation moves data between systems on fixed rules. An AI agent can also read a PDF, a call or an email and decide what to do. Most real jobs need both, which is why the older job title — business process automation consultant — still describes most of the work.
Do you only work with the systems on your list?
No. Salesforce, QuickBooks Online, BILL, Stripe, Airtable, Google Workspace, Slack and Twilio are where we've run agents in production. Anything with an API is fair game.
How do you price process automation consulting?
Fixed fees. Discovery is a fixed price, the build is a fixed project price, and running the agent is a flat monthly fee. No hourly meters.
What if discovery says nothing is worth automating?
Then that is what the document says and the engagement ends there. It happens, usually because the volume is not there yet or because the rules only live in one person's head. You still leave with a map of your own manual work, which is worth having either way.
Who needs to be involved on our side?
The person who actually does the work, for a few hours across the first fortnight, and whoever can grant access to the systems. Not a steering committee. If the only person available is a manager describing the job from memory, we find the exceptions late instead of early, which is the expensive way round.
Do we own what you build?
Yes. The code, the credentials and the documentation are yours, and you can take them and run the agent in-house. Most customers would rather we operated it, which is what the monthly fee covers, but that is a choice rather than a lock-in.
Show us one workflow. We'll show you the agent for it.
A 30-minute walkthrough on your actual systems, not a demo account. No deck.