AI capital allocation
Know what your AI spend is actually buying.
Arbitir connects every AI dollar to a person, team, product and business outcome so budgets can continually allocate toward what works.
A complete governance system
Every dollar across AI providers, API keys, and enterprise seats, in one place, across teams. Live controls to stay on budget.
Attribution to real people
Spend resolved to teams, people, and surfaces. Your organization graphed to maintain span of control.
Spend against output
AI cost per dollar of business outcome. You define measures of output and we enforce capital discipline.
Attribution
Providers bill by account. Budgets belong to people.
No invoice crosses that gap. Arbitir resolves the month to a department, a team, a role, a person and the surface, and queues what it cannot attribute.
- Anthropic, API
- $184,300
- Claude Enterprise, 210 seats
- $96,000
- OpenAI, API
- $121,400
- ChatGPT Enterprise, 169 seats
- $84,500
- Total
- $486,200
- Which department drove the increase?
- not reported
- Which roles concentrate the spend?
- not reported
- What output sits against any of it?
- not reported
- Engineering$142,300
- Sales$96,240
- Unattributed, 4 shared keys$75,800
- Marketing$71,640
- Customer success$58,210
- Support$42,010
- Company$486,200
- Sales$96,240
- Enterprise AE$34,600
- D. Kessler$3,180
- Claude Code$1,940
How it works
Connect once. Improve continuously.
Arbitir creates a recurring capital discipline: observe the input, define the output, govern the decision, and use the result to improve the next allocation.
- 01
Connect
Connect your AI models and business systems. All surfaces report into a single place.
- 02
Understand
See where AI spend is going, who owns it, and what business output sits alongside it.
- 03
Act
Allocate budget toward the work that earns it and set controls by team, role, person, or project.
Approvals
Requests arrive through four channels and are decided in one.
Every route lands in the same queue, each request carrying what the person already consumes and what it returned. The decision takes a click, and writes a ledger row.
hey, hit my Claude limit again. any chance you can bump it before my 9?
who owns the OpenAI budget? need a key for the RFP work this week
Fwd: Re: Re: AI budget for the Q4 pilot (14 in thread)
did this get approved? been blocked since yesterday
can we get 3 more Claude seats for the onboarding cohort?
sorry to message late, still stuck on this one
No amount. No use case. No record.
- D. Kessler Enterprise AE+$150 / mo
Account research before the first call
$310 a month, 1.1x the role median. $1,290 of pipeline per AI dollar, 1.1x.
- M. Novak SDR+$80 / mo
RFP drafting
$96 a month, 0.5x the role median. No outcome recorded yet.
- A. Whitfield Enterprise AE+$400 / mo
Multi-threading the top 20 accounts
$610 a month, 2.1x the role median. $1,836 of pipeline per AI dollar, 1.6x.
Leverage
The same dollar earns a different return in different hands.
Across one role, the return on an identical dollar varies by a multiple. Find the highest return, document the practice, extend it to the rest of the role.
A. Whitfield. 2.1x the median spend, 3.4x the median pipeline: 1.6x the return per dollar.
Account research before the first call
A. Whitfield · Enterprise AE
- 01Pull the last four earnings calls and the risk factors from the 10-K.
- 02Write the three questions the buyer cannot answer from their own website.
- 03Build the agenda from those three questions and nothing else.
They write it themselves. Nothing is read from their account.
Extended to 9 AEs at $610 a month. One ledger row each.
The 9 on the playbook
Peers, same segment, no playbook
Pipeline per AI dollar, 60 days either side.
Medium confidence. A reading, not a cause.
The companies that win the AI era will not be those that spend the most on AI. They will be those that earn the highest marginal return on each incremental dollar of AI capital.
Observe the flows. Define the output. Measure the lift. Move the allocation. Scale what works.
Product
From control to active allocation.
Spend observability and governance builds the ledger. Outcome definitions create the evidence. Allocation turns that evidence into better decisions.
- 01
Controls
Governance policies, budgets, approvals and ledgers
- 02
Insights
Spend against budget and outcomes, utilization and unit economics.
- 03
Outcomes
Success metrics, definition building, and outcome loops.
- 04
Allocation
Recommend and act where AI capital maximizes ROI.
Use cases
Finance owns the budget. Operators own the productivity curve.
The same capital ledger becomes useful when it is translated into the operating outcomes each executive already manages.
- CFO
Finance
Govern the portfolio and improve capital productivity.
- CRO
Sales
Allocate AI behind the motions that create pipeline and revenue.
- CCO
Customer success
Connect AI capacity to retention, expansion, and customer productivity.
- CMO
Marketing
Put AI spend behind programs that create qualified demand and pipeline.
- CTO
Engineering
Connect AI coding capacity to delivery throughput and product work.
We read usage metadata.
Never your prompts.
Arbitir ingests token counts, model names, timestamps, and dollar amounts. It never reads prompt content, completions, or conversation text. There is no code path that could, and that is enforced in the schema rather than promised in a policy.
Credentials you connect are validated once against the live API and then written to a vault. The application database stores the name of the secret, never the secret. Every connection, every credential read, and every change to a budget or a decision is written to an audit log.
Start allocating AI spend toward what works.
We grant access to a select number of companies each month, based on fit. Describe how AI spend is governed today and we will respond within two business days.