Docs Start here Getting Started

Getting Started.

Six steps, in order. Each one is on one screen, and the checklist in the product marks each one done.

1. Connect a source

Paste a read-only key, or sign in on the vendor's page. The key is tested before it is stored. About this screen →

Give each AI cost the team that owns it. About this screen →

Each opportunity shows what it is based on and the change to make. Approve, assign or reject it. About this screen →

Put a monthly limit on a team, agent, model or vendor. About this screen →

Agents: See the agents that make the changes for you, and what each could do now.

Exec summary: See your spend, its owners, what you can save, and your budgets.
Actions: See everything you need to decide or fix, most urgent first.
By department: See every department and its teams by spend, use and cost per result.
Support: See what one support resolution costs, and how often closed conversations reopen.

Connect: See every source, connected or not, and what each one reads.
Credits: See what credit you have left with each vendor, and when it runs out.
Contract rates: See what you pay for each model, at list price or your own rate.
AI register: See every AI tool, agent, MCP server and API key, with its owner, budget, limit and approval.
Seats & plans: See every seat you pay for, its cost, and when it renews.
Additional costs: See the AI costs you entered, such as your own GPUs or training runs.
Spend: See what your AI spend was and what changed against the period before.
Dashboards: See your own pages of the figures you check most.
People: See each person's AI use: spend by week, models, the prompt cache, failed calls, their cap, and the coach's last note.

Attribution: See who owns each dollar and what work it paid for.
Org chart: See where each team sits, in its department and business unit.
Your usage: See your own AI use, and the last note the weekly developer coach sent you.
Shared costs: See how shared costs no one owns are divided, and by what method.
Outcomes: See what your AI spend produced, and what each result cost.
Cost per outcome: See what one successful outcome costs in each workflow.
Value: See if one workflow's AI is worth it, with an owner and review date.

Benchmarks: See how your AI spend compares, and what closes your biggest gaps.
Every way to save: See every known way to cut AI cost, and where you stand on each.
Open models: See what your tokens would cost on open models such as Qwen or Llama.
Opportunities: See every opportunity found, with its evidence and the change to make.
Scenarios: See what a change would have saved on last month's AI calls.
Ledger: See every decision on an opportunity, with the saving expected and measured.
Engineering: See whether AI is helping each team: its pull requests against a person's, matched fairly, and the size at which AI code comes back.
Cost: See what a shipped change costs at what you are billed, seats included, and where money leaks.
Seats: See which plan each developer should be on, who hits a limit, and how to get more from each seat.
Changes: See the changes your engineering numbers point to, each with its evidence, the exact setting and how it will be checked.
Developers: See how coding sessions end, and try coding reads on one session file before turning it on.

Budgets: See every budget, on any team, product or vendor, and where its month lands.
Forecast: See where this month lands, with a range and the assumptions behind it.
Spend rules: See policies beyond monthly limits, such as a retry cap or error-rate guard.
Approvals: See the AI calls that would have waited for approval, and who decides.
Compliance: See every AI tool in use, whether it is allowed, who holds it, how its checks stand, and the asks for new ones.
Access: See which teams and people hold which AI tools, and the seats nobody uses.
Regulations: See how ready you are for each AI rule that applies to you, every gap with how to close it, and your AI incidents.
Alerts: See every alert raised this period, its cause and who acted on it.
Audit trail: See every change in your organization, newest first.

Ask Valistry: See the answer to a question in plain words, with its figure and evidence.

People: See who is in your organization, their role, and what each one sees.
Plan & billing: See your plan, what it costs and what it includes.
Notifications: See what Valistry sends you, how often, and where it goes.
SSO: See how your identity provider signs people in and manages access, on Enterprise.
Nightly export: Send your figures to your own endpoint each night, on Enterprise. Files are on each page, under its share icon.
Access: See what Valistry reads and writes at each level, Light, Medium and Deep, and make the keys each needs.

Terminology: Every term on a screen, in a mail or in an answer, defined in one place.
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 works out on its own, how to connect it.
Amazon Bedrock guide: What we read from an AWS estate, the four changes that cut its bill, and what we cannot yet see.
Compliance: The AI inventory and the decision record this produces anyway, and which obligation each speaks to.
Developers: The same figures as JSON, and a way to import rows from a system we have no connector for.