Work with us
AI that works on real company data.
For US distributors and manufacturers without an AI team of their own. We find where AI pays off in your operations, build it on your own systems and data, and keep it working.
01 / 04
Why most AI projects stall
The model is rarely the problem. Projects stall on scattered data, unclear owners and no way to check the answers.
That is where we start: with your data and one process, before any tool.
- Data spread across systemsOrders in email, prices in a spreadsheet, stock in the ERP. AI can only be as right as the data it reads.
- Nobody owns the processA tool is bought, but no one is in charge of the job it was meant to change, so nothing changes.
- No way to check the answersA good demo is not proof. Without tests on your own documents, nobody knows when it is wrong.
- Built for everything at onceThe projects that work start with one process, one number to move and one person in charge.
02 / 04
Three ways to work with us
Most clients start with the assessment. It is the quickest way to know whether AI is worth it for you, and it leaves you with a plan you own either way.
- 01
AI assessment and roadmap
Two to three weeks, fixed scope
We map your processes and where your data lives, find the jobs AI can take on safely, and estimate the hours it would give back and what it would cost.
- A map of your processes and data
- Opportunities ranked by value and risk
- A one-page roadmap for the year
- A first project, scoped and ready to start
- 02
Implementation
Per project, scoped from the roadmap
We build it end to end on your own systems: your ERP, your inbox, your documents. Anything the system is unsure of goes to a person, with the reason.
- A working system on your data
- Tests on your own examples, run before every change
- Logs of everything it did, and a way to undo it
- Handover your team can run
- 03
Fractional Head of AI
Monthly, for companies without an AI lead
We act as your AI lead: we own the roadmap, run the builds, watch quality and cost, and keep everything written down, so AI keeps paying for itself.
- Your AI roadmap, kept current
- Builds and improvements each month
- Quality and cost checked, with a monthly review
- One named lead and a client portal
03 / 04
Typical first projects
For distributors and manufacturers, the first project is usually a job your team does by hand every day, where the time saved is easy to count and a person can check anything unsure.
- Orders from email and PDFRead, checked against your customers, products and prices, and entered as drafts.
- Supplier bills matched to ordersEach bill matched to its purchase order and sent to the right person to approve.
- Quotes from your price listA first draft from the request and your prices, checked by your salesperson.
- Plain-English questions on your ERP dataAnswers from a safe, read-only layer, tested against questions with known answers.
Every project runs the way we describe on How we work: a client portal, every task written down, checks before anything goes live, and tested backups.
04 / 04
How it starts
No pitch deck. A conversation about your operations, and an honest view of where AI fits.
- First
A 30-minute call
Your processes, your systems, and where the hours go each week.
- Then
A straight answer
Whether AI is worth it for you yet, and where we would start. Sometimes the answer is to fix the data first.
- If it fits
A fixed-scope assessment
A written proposal for the assessment, with what you get and when.
Tell us where the hours go.
The first call is free. You'll leave with a straight answer on whether AI is worth it for you yet.