Five ways we help. Most people start with the first one.
Everything below exists because someone asked for it after an audit. Nothing below requires you to understand how a language model works.
The AI audit
Two weeks, one short report, a fixed $495. What to automate first, what to leave alone, and what each line would cost.
We sit with the people who do the work, not just the people who describe it. We map where time actually goes, where the copying and re-typing happens, and which of it a model can do reliably today. You get a plain-English report with a ranked list and a price next to every item. No slide deck.
What you get
- A ranked list of what to fix, in order
- A price on every line, so there are no surprises
- A list of what not to automate, and why
- A walkthrough call to argue with us about it
Good for anyone who has been told to "look into AI" and would like to be done looking.
Book the audit, $495
Workflow optimization
We take the one process that eats your week and rebuild it with AI doing the boring middle and a person signing off at the end.
Quotes, intake, invoicing, scheduling, donor follow-ups, the monthly report. The pattern is always the same: a form, a spreadsheet, three emails, and somebody retyping it all into another system. We cut the retyping, keep the judgment calls human, and make sure it fails loudly instead of quietly.
What you get
- One workflow rebuilt end to end
- Connected to the tools you already pay for
- A human checkpoint wherever money or a customer is involved
- A one-page runbook for when something looks wrong
Good for teams that know exactly which Tuesday afternoon they want back.
Talk about a workflow
Custom tools
When the off-the-shelf thing almost works, we build the thing that does. Small, boring, and yours.
Sometimes the gap between what a tool does and what you need is the whole problem. We build small internal tools: a quoting engine that knows your pricing rules, a report that writes itself from your data, an intake portal your customers can actually use. Built to be maintained by someone other than us.
What you get
- A working tool with your data, not a prototype
- Hosted where you want it, documented so you can leave us
- Reliability engineering baked in, so it survives real use
- Ownership of the code
Good for businesses with a process that's specific enough to be a competitive advantage.
Describe the tool
Training and adoption
The tool is the easy part. We teach the team to use it, write the rules for what AI may and may not touch, and stick around until nobody needs us.
Half of AI projects fail because nobody told the team how to use the thing, or what it's allowed to see. We run short, practical sessions with real work from your inbox, write an AI use policy a normal person can read, and set up office hours for the first month.
What you get
- Hands-on sessions built around your real work
- A one-page AI use policy for the team
- A month of office hours
- A short list of the tools worth paying for, and the ones to cancel
Good for the eleven-to-two-hundred-person company where three people are quietly doing all the AI.
Set up training
Ongoing optimization
Models change every month. We keep what we built working, cheaper, and a little better, so you don't have to read the release notes.
A monthly retainer. We watch the tools we set up, swap models when a better or cheaper one ships, tune the prompts as your business changes, and take the next item off the audit list when you're ready.
What you get
- Monitoring of everything we built
- Model and cost tuning as the market moves
- A monthly note on what changed and what it saved
- Priority on the next item from your list
Good for anyone who liked the first project and would like the second one without a kickoff meeting.
Ask about a retainerNot sure which one you need? That's what the audit is for.
Two weeks, a short list, and a price next to everything.
Credited toward any project we start within 60 days.