The approach
Three articles. No annexes.
Most AI consulting produces documents about what you could do. Ours produces tools your team uses, and owns. The engagement has three parts, and each one ends with something in your hands.
Discovery
Two weeks
We find where AI actually pays.
We spend two weeks inside your operation, with the people doing the work, not just the org chart. We map the processes that consume the most hours, the decisions that wait on information, and the knowledge locked in inboxes and heads.
You get a short, costed plan: the two or three applications of AI that will genuinely earn their keep in your business, what they will save or unlock, and what we recommend you ignore no matter what the conference keynote said.
You receive
- Process map of AI-ready work
- Costed shortlist, ranked by payback
- Sovereignty assessment: what must stay in-house
- A plan a board can approve
Implementation
Six to twelve weeks
We build working tools on ground you own.
Open-source models, self-hosted or in your private cloud. We build the first tool end to end: connected to your real systems, tested on your real work, used by your real people before we call it done.
Your team is in the room from the first week. Delivery draws on the engineering bench of Agiledrop: fifteen years of shipping software for organizations that cannot afford it broken.
You receive
- Working software in production
- Running on your infrastructure
- No proprietary dependency you cannot replace
- Documentation your team actually reads
Handover
For as long as you want us
Your people run it. Your keys, your models, your data.
The finish line is your independence: your team operates, extends, and owns the stack. We train the people who will run it and leave behind the playbooks to do so.
Then we stay exactly as long as you find us useful, and not a week longer. If it does not run without us, we have not finished.
You receive
- Trained internal owners
- Operating playbooks
- A stack you can hire for on the open market
- An exit that costs you nothing
Start with the two weeks.
Discovery is fixed-scope and fixed-price. If we find nothing worth building, we will tell you that too, in writing.