// THE ROUTING ASSESSMENT
One week. One workflow. Proof of what AI should cost you — $4,500, fully credited if you continue.
Before you commit to a platform, a model, or a hire: we take one workflow your team does by hand, run it across 15+ AI models, and hand you a report — what it would cost, which model is good enough, and what must stay human.
1 workflow
Fully measured, end to end
$0
Platform commitment required
5 days
Kickoff to written report
"Unlimited users — your approvers, auditors, and reviewers are free. You pay for the work agents actually do."
That's how everything we deploy is priced. No per-seat licenses, no charge for the people who check the work.
// what you get
A report with numbers, not adjectives
Sample pages from a delivered assessment. Client data redacted — the structure is exactly what you'll receive.
The routing table
| Task type | Model | $/1k |
|---|---|---|
| small | 3.20 | |
| mid-tier | 7.10 | |
| mid-tier | 6.50 | |
| nano | 0.80 |
Every task type in your workflow, mapped to the cheapest model that holds quality — with the method noted per row.
The cost/quality frontier
Every model we tested, plotted by measured quality against cost per 1,000 tasks. The circled point is your routing pick.
The savings summary
−77%
Projected monthly cost at your volume, against your own baseline — model spend or staff hours.
// run your numbers
What would routing save you?
What's your baseline?
- Manual cost today
- $6,400/mo
- Routed pick: small (quality 91)
- $4.80/mo
- Human review (15% of tasks)
- $960/mo
- Estimated saving
- $5,435/mo
Quality bar for claims-status summarization: 90. The routed pick is the cheapest model that clears it in our benchmark.
These are estimates from our benchmark data. The assessment gives you your numbers.
Book the assessmentNo procurement needed
$4,500, fixed fee, one week. Sized for a discretionary budget — no RFP, no committee, no platform contract.
Runs on SectorFlow One
Every assessment runs on the same governed platform we deploy for clients — so the assessment doubles as a working product demo.
Credited toward the Build
Continue to a deeper Assessment or the Build and the full $4,500 comes off the price. Stop after the assessment and you keep the report. Either way, you got the numbers.
Want the full engagement ladder? See pricing
// questions
Asked before every assessment
We don't use AI at all yet. Is an assessment premature?
No — it's the cheapest possible first step. Instead of committing to a platform or a pilot, you get proof on one real workflow: what it would cost, which model is good enough, and which parts must stay with your team. You decide what to do next from evidence, not a vendor pitch.
What do you need from us?
One workflow your team does by hand, 20–50 redacted samples of its inputs and outputs, and a one-hour kickoff call. No AI infrastructure, no API keys, no IT project.
What if the assessment says AI isn't worth it for this workflow?
Then the report says so, with the numbers that prove it — and you just avoided a five-figure pilot that would have found out the slow way. The report is yours either way.
Is our data safe?
We work from redacted samples, sign an NDA on request, and run everything on SectorFlow One — with audit logs, approval workflows, and tool-permission scoping built in. The same governance posture we'd deploy for you.
What does the $4,500 buy, exactly?
A written report: the routing table for your workflow, a cost/quality chart across 15+ models, a savings estimate against your current manual cost, and a human-in-the-loop map. Fully credited toward a deeper Assessment or the Build if you continue.
// book the assessment
Pick a kickoff slot
One hour to scope the workflow. Written report back within five business days. $4,500, fully credited toward a deeper Assessment or the Build.