Valistry

Docs Start here Terminology

Terminology.

Every term on a screen, in a mail or in an answer, in one place. A word here means one thing everywhere; a word not here is not used.

Every term, defined

Estate
Every AI source, model, agent, seat and workload an organization pays for.
Ledger
The reconciled record of every AI dollar, attributed, with what it produced.
Source
A connector that reads a vendor: OpenAI, Anthropic, Bedrock, a gateway, a seat tool, a business system.
Bill
What the vendor charged, read from the vendor, shown beside the ledger, never used to fill it.
Reconciled
The ledger's total for a source against the bill for the same days. Exact, or the difference is shown.
No owner yet
Spend no rule, tag or trace has given an owner. A named row on every screen, never hidden.
Allocation
A decision dividing shared cost across teams or workflows, by a method the business chose. Never evidence, never attribution, and it never moves the total.
Teams
Team inside department inside business unit. Each AI call keeps only its team, so a reorganization changes no past records.
Department
A group of teams, such as Engineering, Finance or Support.
Coverage
The share of spend with an owner. It rises when you publish a rule, never by estimate.
Opportunity
A saving or a risk Valistry found in your AI calls, with its evidence, the AI calls behind it, the change to make and the work it needs.
The work a change needs
Low, medium or high effort. Low effort needs no code: a rule, a cap, a flag or a seat.
Prompt signature
A hash of a prompt's shape, never its text. Calls sharing one are the same prompt doing the same job.
Low-effort savings
The opportunities that need no code, summed. The one figure Starter shows of the opportunities.
Outcome
The business result a workflow produced, from the system that recorded it, with its evidence class.
Evidence class
Observed, declared or unmeasured. Every cost-per-outcome figure says which.
Cost per successful outcome
Attributed spend divided by successful outcomes, retry waste kept apart. Never cost per token.
Ledger
Every decision on an opportunity: expected on approval, verified after the change, against the same baseline.
Verified saving
What the ledger measured after a change, at least seven days on, against the same baseline. Expected is never overwritten. A change taken back stops counting, and the measurement stays as history.
Realization rate
Verified savings over expected savings. The company's own metric.
Lever
One way to cut AI cost, such as prompt caching or a cap on retries. Every lever lists them all, in eight groups.
What good looks like
A measure beside your own best month, published figures, and the median of companies your size once five or more share.
Budget
A monthly limit on a team, an agent, a product, a customer, a model or a vendor — in dollars or tokens — set from what that thing already spends.
Policy
A rule with what happens if it goes over: observe, warn, require approval, enforce. Tried on your past spending before it is live.
Governed spend
The share of AI spend under a budget with an owner.
Alert
Something one of the seven watches raised: a spike with its cause, a budget, a policy, approvals waiting, a problem back after a fix. Sent once; open until someone acknowledges or dismisses it.
Kind of work
What the AI calls were for: coding assistant, chat assistant, in-product agent, retrieval, speech, image, evaluation, tool calls.
Priced by
What the vendor priced: tokens, seats, audio seconds, images, the bill itself, or use a plan already covers.
View
A named part of the estate with its own channel. It narrows a question and never re-answers it; it is a convenience, not a permission.
Rate agreement
What you negotiated with a provider, entered because no API reports it. It re-prices this period and never touches the bill.
Starter, Platform, Enterprise
Free at any spend; priced case by case; the Platform tier on an order form above $100,000 a month.

Where to next

What we collect
What one line of the ledger carries, and the seven things it can never hold.
Connectors
For each source: what it reads, the access it needs, what it attributes on its own, how to connect it.
The read API
The same figures the screens show, as JSON, with the key or the sign-in the editor uses.