4
Consulting
You bring the AI problem. We answer in writing with a scope and a fixed quote, and we build it with the same systems we research.
What we take on
- Agents that remember
- Assistants and coding agents that keep what they learn from one session to the next, recall it by meaning, and can be corrected.
- Private models on your own machines
- Local language models for work that shouldn't leave the building, sized to the hardware you have.
- AI inside the systems you already run
- Claude or other models added to your CRM, portal, store or internal tools, through tools the model can call and you can audit.
- Checking AI claims
- Does the feature do what the demo said? We trace the claim to its records and report what holds and what doesn't.
How it works
- Write to us. What you're trying to do, what you have now, and what's in the way.
- Get a written scope and a fixed quote. What we'll build, what we won't, and how you'll know it works.
- Decide. Nothing starts until you say yes.
- We build and show the evidence. Working software, with the tests and the numbers behind it.
Everything happens in writing: email and a shared project page. No phone or video calls.