Four phases. A written sign-off at every gate.
You can stop at any of them.
No open-ended retainers and no black boxes. Each phase ends with something concrete in your hands and a written approval before the next one starts — so you always know exactly what you're buying, and exactly what you own.
No pitch, no obligation, NDA-friendly.
What happens, what you receive, and the gate before the next step.
Workflow teardown & AI strategy
We sit with your team and watch how the work actually gets done — what's public vs. confidential, where the friction is, and which tasks repeat. No tools are touched in this phase.
- A written AI strategy memo in plain English
- A lane map: every workflow we reviewed, sorted into Lane 1, 2, or 3
- A tool recommendation with the reasoning spelled out
- A first draft of your do's-and-don'ts rules
Sign-off gateYou approve the strategy and lane map in writing before we configure anything.
Approved tools, set up right
We configure the approved assistants for your team under enterprise terms — training disabled, access controlled, rules attached. One approved stack instead of five secret personal accounts.
- Configured enterprise AI accounts, training disabled
- Plain-English usage rules your team can actually follow
- A prompt library built from your real recurring tasks
- A kickoff training session for everyone who'll use it
Sign-off gateYour compliance, legal, or IT lead signs off on the configuration before rollout.
First workflow, built and piloted
We pick one recurring task from the lane map and turn it into a reviewed process: approved sources in, a human check in the middle, a saved record at the end. Then we pilot it with the people who own that work.
- One live workflow with a human-review step built in
- A reviewer checklist so the check is real, not a rubber stamp
- Saved-run records from day one
- A pilot report: what worked, what we changed, what's next
Sign-off gateYou review the pilot and decide — extend it, adjust it, or stop. Your call.
Train, hand over, maintain
We train your people by role, document everything we built, and hand over the keys. Ongoing maintenance is optional — quarterly tool reviews and refreshers — but the goal is your independence, not our retainer.
- Role-based training with examples from your own work
- Full documentation: tools, prompts, workflows, records
- A handover package — everything belongs to you
- Optional: quarterly reviews as models and vendors change
Sign-off gateHandover sign-off. You own it all; nothing is held back.
Everything we build belongs to you. Nothing is rented.
Every document
Strategy memos, rules, prompt libraries, training playbooks — written for your team, left with your team.
Every workflow
The processes we build run in your accounts, on sources you approved, with your people reviewing.
Every record
Saved runs, sign-offs, and review trails stay in your environment. Compliance answers from a package, not from memory.
Your independence
The setup is model-agnostic. Change vendors, change models, or part ways with us — nothing needs rebuilding.
Why there are no prices or durations on this page
Because they'd be made up. How long each phase takes depends on your team's availability and your approval process — a fund with a compliance review cycle moves differently than a ten-person operating company. Quoting a number on a website would tell you nothing about your engagement.
What we commit to instead: the sequence above, a written scope before we start, and a sign-off gate at every phase — so the spend is never open-ended and stopping is always a clean option. Book a call and we'll scope yours in writing.
Want this scoped for your team?
A 20-minute call. You describe how the work gets done today; we tell you what phase one would cover and put the scope in writing.
NDA-friendly. No obligation. Or email hello@cleverfoxailabs.com