For Finance

"AI spend" shouldn't be
a line item you can't explain.

Most AI tools price by seat and bill by surprise. Here, the plan price is published, every message records its cost in dollars, and budget caps are hard stops — not notifications. Five questions finance asks, answered with numbers you can check.

What you'll want to know

Cost, return,
and the ceiling.

The short version: pricing is public, spend is metered per message, caps are enforced at runtime, and the pilot is measured against metrics you set before it starts.

01

What does it actually cost?

Per-plan pricing is published, in full, on the pricing page — no call required to learn the number. Underneath the plan, every message an agent produces records its token count and USD cost, so the unit economics aren't a quarterly estimate: they're a column you can sum, per worker, per workspace, per day.

See the plans
02

How do we know it's paying off?

Analytics scorecards show what each worker did — tasks completed, cost per run, error and escalation rates — against the time those tasks would otherwise consume. Before you commit, the ROI calculator lets you model the same math with your own headcount and task volumes, so the business case is yours, not ours.

Model your numbers
03

How do we stop runaway spend?

Hard daily and monthly budget caps — hard as in the spend stops, not as in someone gets an email. Breach-date forecasting warns you before a cap is hit, and circuit breakers auto-pause a worker when a cost or error threshold trips. There is no configuration in which an agent quietly burns through a month's budget overnight.

How caps are enforced
04

What's the procurement process?

Standard terms, a DPA on request, and a structured 30-day pilot framework with success metrics and exit criteria defined up front — the whole evaluation path is written down on the procurement page. Security questionnaires get completed as part of the job, not as a favor.

The procurement one-pager
05

How fast is time-to-value?

Solution packs install a complete working department — agents, tables, knowledge, policies — in about 15 minutes. That means the pilot spends its 30 days measuring outcomes, not building infrastructure, and you can trial the actual workflow before committing budget to it.

Browse the packs
What to take with you

Three things to bring to the budget meeting.

Each one is built to survive scrutiny from someone who didn't attend the demo.

Cost visibility lives in Analytics; the controls that keep spend inside the line live in Governance. Security answers for your counterparts are on /trust.

For your evaluation

Get a cost model for your team.

Tell us your team size and the workflows you're considering, and we'll send back a plan recommendation with the budget-cap configuration to match.

Request the model