I automate one workflow in your business, end to end, in 4 weeks.
Fixed price — $10,000 CAD. Not a retainer, not a discovery phase, not a pilot that never ends. One workflow your staff does by hand today, running by itself when I leave.
You already know which task it is
It is the one somebody on your team does every single day, by hand, that produces nothing a customer ever sees. Retyping intake forms. Chasing a schedule across three systems. Reading invoices into a spreadsheet. Writing the same summary from the same six fields. Reconciling what one system says against what another one says.
You have probably been told this can be fixed with AI. What you have likely been offered is a strategy deck, a year-long transformation programme, or a chatbot demo that impressed nobody. Meanwhile the person is still retyping the forms.
I do the opposite. One workflow. Four weeks. A fixed number agreed up front. It either runs by itself at the end or you have lost nothing but the cost of the First Phase.
How it works
Two phases. The first is small and self-contained, so you can find out whether I am any good before you commit to a build — and nothing in it is wasted if you go ahead.
Discovery & Audit
- Built on Claude in your account, under your keys, with your data staying yours
- A half-day walkthrough of how work actually moves through your business
- Every manual workflow written down, with hours-per-week attached to each
- A ranked shortlist: what is worth automating, what is not, and why
- A written document you keep — and can hand to anyone, including a competitor of mine
The Sprint
- Week 1 — Business discovery and an audit of the manual operations worth automating. Skipped entirely if we already did the First Phase.
- Week 2 — I build against your real data, not a sample. You see it working by the end of the week.
- Week 3 — Your staff use it while I watch, and I fix what breaks. This is where most projects skip and fail.
- Week 4 — It runs in production. Handover, written runbook, and training for whoever owns it after me.
$7,500 over 3 weeks if we have already done the Discovery & Audit — Week 1 comes out of the scope and out of the price.
Fixed price either way. If it runs long, that is my problem and my cost, not a change order.
What is in, and what is not
A fixed price is only honest if the edges are drawn. Here they are.
In scope
- One workflow, defined in writing before a dollar moves
- Integration with the systems you already run
- Reading and writing your real documents, forms and records
- Error handling, and a path for a human to step in
- Written runbook plus a training session
- The code, in your repository, under your ownership
Not in scope
- Replacing your core system of record
- A second or third workflow — that is a second sprint
- Ongoing support after handover — that is a retainer, priced separately
- Anything needing a certification I do not hold
- Custom hardware, or on-premise infrastructure I cannot reach
- A promise that AI will let you cut staff. I do not sell that
Who is doing the work
Me. Not a junior, not an offshore team, not a project manager who relays questions to someone you never meet. If we work together, the person on the call is the person writing the code.
I have spent 25 years as a software architect building production systems — including at 3M Canada, Union Gas (Enbridge), the University of Ottawa, The Weather Network, Voices.com (a marketplace with over a million members), and a banking platform used by family offices. Six Sigma Green Belt. I have also been the person on the other side of this: the one who inherits the automation somebody else built and abandoned. That is why the runbook and the handover are in the fixed price rather than sold back to you later.
You may not be paying for all of this yourself
Training your staff is one of the most heavily subsidised things a Canadian employer can buy. If we structure the sprint so your people are trained on the system as it is built — which is how I run it anyway — a large share of the cost may be recoverable:
- BC Employer Training Grant — covers 80% of eligible training costs, up to $10,000 per employee. Budgets typically run out by late summer, so the window matters.
- Canada–Alberta Job Grant — government covers two thirds, up to $10,000 per trainee per year, to $100,000 per employer per fiscal year
- Canada–Alberta Productivity Grant — half the cost of eligible team training, up to $5,000 per employee per year
- BDC LIFT and SR&ED tax credits — for the build itself, where it involves genuine technical uncertainty
Canada has a stated national target of lifting business AI adoption from roughly 12% to 60% by 2034. The money exists because the country has decided it needs you to do this.
Fair questions
Four weeks is not long enough to do anything real.
For one workflow, with the scope written down and no committee in the way, it is. The reason software projects take nine months is rarely the building — it is the deciding. The First Phase does the deciding first, on paper, for a quarter of the price.
What if it does not work?
Then you stopped after the First Phase and you keep the written analysis of your own operations — including the parts that say do not automate this yet. I would rather lose a sprint than sell you one that fails: I am building a reference list here, and one bad outcome in a market this size costs me more than one fee is worth.
We have no technical staff to maintain it.
That is the normal case, and it is why the runbook and training are inside the fixed price rather than an upsell. If you want me to keep an eye on it afterwards, that is a separate monthly arrangement and it is optional. It genuinely is optional.
Our data is confidential — patient records, client files.
It stays in your account, under your keys. I will sign an NDA before the audit, not after. If your sector has rules about where data may live, tell me at the audit stage and I will tell you honestly whether I can build it inside them.
Why should we use you rather than a big firm?
If you want a transformation strategy across the whole organisation, use a big firm — that is a real service and I do not offer it. If you want one specific task to stop eating a person's week, a partner at a large firm bills more per hour than this entire sprint costs.
Can you just tell us what to buy instead?
Often, yes, and sometimes that is the honest answer — if an off-the-shelf product already does it, the audit will say so and name the product. You will have paid a small fee to avoid a large mistake.
Start with the audit
$2,500 and 5 business days. You get a written analysis of where the manual work actually is in your business, and you keep it whatever you decide next. Go ahead and it becomes Week 1 of the build, already done — so the sprint runs 3 weeks at $7,500 rather than 4 weeks at $10,000.
Taking only 2-3 sprint clients per quarter. That is a capacity statement, not a sales tactic.