Opportunity review
Which workflows benefit from AI, which don’t, and how we’ll measure the difference.
AI implementation
Useful AI.Clear boundaries.
AI makes impressive demos easy. Making it dependable is the hard part. I build AI workflows and tools that fit real processes, with defined permissions, human review where it matters, and costs and quality you can see. I’ve been building AI products since before they were the obvious thing to build.
Sound familiar?
You probably need this if…
- Your team uses AI tools informally and nobody knows what data goes where.
- You built an AI prototype that impressed people but never reached production.
- Drafting, sorting, or summarizing work eats hours every week.
- You’re not sure which AI ideas are worth the cost and risk.
What you get
Decisions, delivery, and someone who stays.
Honestly, you don’t need me if
- You want AI added because competitors mention it, without a process to improve.
- The work needs guaranteed-correct output with no human review and no margin for error.
AI workflows and tools
Drafting, classification, extraction, and assistants, connected to your data and tools.
Boundaries and review
What the AI may read, draft, or do, where a person approves, and how to recover from mistakes.
Cost and quality tracking
Usage, error rates, and cost per completed outcome, visible to you.
Work behind the words
Built, shipped, and still running.
Questions
Straight answers.
Which AI models do you use?
Whichever fits the job, its data rules, and its budget. The model is the easy part to swap; the workflow, permissions, and review points are what make it work.
Is our data safe?
We decide up front what the AI can access and where data is processed, then test permissions with real roles before launch.
What does it cost?
Ongoing partnerships typically run $10K–$20K per month. A focused first engagement is scoped and priced separately, before any work begins. The first call is free.
Who owns the code and accounts?
You do. Accounts, code, and documentation are set up so your business controls them. The handover covers where things run, who has access, and how to stop or change them.
Read before we talk
The thinking behind the work.
Research reportUseful AI. Clear boundaries.
Design the permissions, review points, and recovery paths that make AI useful in a real business.
Research reportThe business of making it workThe distance between an impressive demo and a system your business can depend on.
Research reportThe invisible workdayFind the copying, chasing, and broken handoffs hiding inside a busy working day.
Also available
One partner across the stack.
Most projects touch more than one of these. You get one accountable person for all of them.
Let’s find the first useful move
Tell me what’s getting stuck.
Book a call with IanThe first call is free. We’ll decide what makes sense from there.