Less copy-paste. More actual work.
I build AI tools and automations that take the repetitive work off your team: orders, enquiries, paperwork, product data and reports. They plug into the systems you already run, and a person stays in charge of anything that matters.
Signs a task should be automated.
The same job, every day
Someone copies orders, leads or invoices from one screen into another. It's nobody's favourite hour.
The same five questions
Shipping times, opening hours, where's my order. Answered by hand, again, in three different inboxes.
A catalogue nobody can keep up with
Hundreds of products missing descriptions, tags, alt text or the French version.
Numbers live in six tabs
Monday starts with exporting reports from every tool and pasting them into one spreadsheet.
Things fall through the cracks
A lead sits unanswered over the weekend. A low-stock warning gets noticed after it sells out.
You tried ChatGPT, now what?
It's useful in a chat window, but it doesn't know your products, policies or customers, and it doesn't touch your systems.
Software does the steps. You make the call.
A typical job: wholesale orders arrive by email as PDFs, and someone retypes each one into Shopify. Now the order is read, matched and checked, and lands as a draft with a reply ready. Your team checks it and clicks approve.
- We start with one task you repeat every week
- Anything a customer sees gets your OK first
- Runs in accounts you own, with an off switch
- Monthly running cost estimated before I build
- Email from a wholesale buyerPO-2291.pdf attached09:14
- Read the PDF12 lines, delivery Thu09:14
- Matched every SKU in Shopify1 swapped for its new code09:14
- Checked stockall 12 in the warehouse09:15
- Drafted the order and a reply09:15
- Waiting for your OKApprove, edit or reject
Practical AI, wired into your systems.
No partner deals and no favourite platform. The right tool is the one your team can understand and afford to run.
Workflow automation
Triggers, schedules and hand-offs between your tools, on n8n, Zapier, Make or plain code.
AI assistants
Support and internal helpers that answer from your own documents and data, with a person in the loop.
Document processing
Invoices, POs, intake forms and contracts read, checked and filed into the right system.
Content at scale
Product copy, alt text, SEO fields and English/French translations, in your voice and reviewed.
Integrations & APIs
Custom connections where no ready-made app exists: ERP, POS, CRM, accounting and 3PL.
Reports & alerts
Weekly summaries in plain English, and a message when something actually needs you.
AI in your store
Search that understands what shoppers mean, smarter filters and product recommendations.
Internal tools
Small apps and dashboards for your team, so the work stops living in spreadsheets.
Not just for online stores.
If work moves between an inbox, a spreadsheet and another system, there's usually something to automate.
Shops & e-commerce
- ✺Wholesale orders from email to Shopify
- ✺Product data and translations in bulk
- ✺Low-stock and reorder alerts
Offices & professional services
- ✺Leads sorted and followed up
- ✺Intake forms into your CRM
- ✺Quotes and invoices drafted
Clinics, studios & trades
- ✺Booking questions answered
- ✺Reminders and review requests
- ✺Job notes turned into invoices
Nonprofits & enterprise teams
- ✺Donor and member data kept in sync
- ✺Grant and report drafts from your records
- ✺Private AI on your own cloud account
Start with one task.
Automation check
A call and a look at how the work moves today. You get a short written list of the tasks worth automating, with rough costs.
First automation
One task, fixed price, usually live within a couple of weeks. You see it run on real data before it goes live.
Keep going
Add the next task when the first one has earned its keep, or take a monthly block of hours.
Fair questions.
AI gets oversold. Here's how I handle the parts people worry about.
Do I need to change the tools I use?
Usually not. The point is to connect what you already run. If a tool is the actual problem, I'll say so and explain the options.
Where does my data go?
Only to services you've agreed to, under accounts in your name. Business AI APIs don't train on your data by default, and for sensitive work we can keep everything inside your own cloud.
What if the AI gets something wrong?
It will, sometimes. That's why anything a customer sees, or anything that moves money, waits for a person to approve it. Every run is logged so mistakes are easy to spot and fix.
What does it cost to run?
Often a few dollars to a few tens of dollars a month in API and tool fees, depending on volume. You get an estimate before I build, and usage alerts after.
Zapier, n8n or custom code?
Whichever is simplest to own. A ready-made tool is fine for straightforward jobs. Custom code makes sense when the volume, the logic or the privacy needs outgrow it.
Which task would you hand off first?
Tell me about one job your team repeats every week. I'll reply with how I'd automate it and roughly what it would cost.