Explore Flowstate

The Flowstate platform
Flowstate Helm

Know what AI costs. See what you get back.

Understand spending, share budgets and connect AI use to the work it supports.

Recent insights All insights

Buy tokens, rent GPUs, or own the rack?

Open-weight models are now good enough to make self-hosting a financial question rather than a capability one. I modelled four options, an office rack, colocation, rented GPUs and per-token APIs, to find where each one actually wins. The answer turns on utilisation, and it isn't the one the internet keeps giving.

Turn a $3m AI bill into $1.9m

Most teams overpay for AI twice without noticing: they run everything through the default flagship model, and they buy the tokens on a metered plan they can't see into. An intelligent proxy fixes both. Here is the maths, a calculator you can drag, and where Flowstate fits.

Start with the work All use cases

Understand AI spending and value

See where the budget goes and what the business gets back.

Share one AI budget across your tools

Fund people, teams and projects across connected services.

Give your team more capacity

Reduce the preparation, checking and follow-up around recurring work.

Connect AI use to delivered work

Follow AI activity into projects, code and business results.
Quick links
Flowstate / People + AI Talk through your priorities
Documentation
Get help
Docs / Get started

Track AI adoption

Know which teams have made AI part of how they work, who could help others get started, and which paid seats sit idle. Use it when you roll a tool out to more teams, whether that’s engineers taking up a coding assistant or a support team using AI to answer customers, and before you renew seats.

When you’ve finished this checklist, Insights → Agent insights → Adoption shows how many people used AI in the last 30 days, how deeply they use it, which teams are behind, and the seats nobody has opened.

An AI or engineering lead owns this checklist, with a finance lead for seats, IT and security for Flowstate Desktop, and a Flowstate admin.

How adoption is counted

  • Someone is active when they’ve run at least one AI session in the last 30 days. Holding a seat doesn’t count.
  • Sessions come from Flowstate Desktop. Until the apps are on people’s computers, Adoption shows nobody as active, and the people your providers bill for show as holding idle seats.
  • Adoption counts employees on teams. Contractors aren’t counted, except when you group the table by Geography or Employment type.

Before you start

1. Switch on AI spend

What: Ask for AI spend. Who: Implementation lead. Where: Your Flowstate contact. Done when: Insights → Agent insights opens with an Adoption tab, rather than “Agent insights is not enabled”.

2. Put everyone on a team, with their work email

What: Make sure every employee who uses AI is in Flowstate, allocated to a team, with the work email they use at each AI provider and in your device management tool. Who: HR or people data owner. Where: Load your people and teams Done when: The table on Adoption lists your teams, rather than saying adoption is measured against team allocations.

Flowstate matches AI use to people by email. Someone who isn’t on a team isn’t counted in any team’s adoption.

3. Connect the AI providers you pay for

What: Connect each provider’s billing, so the seats you pay for arrive against the people who hold them, where the provider reports that. Who: Flowstate admin, with each provider’s admin. Where: Connect your AI providers Done when: Every provider on your list has a key showing Working under Connected keys.

4. Roll out Flowstate Desktop

What: Put Flowstate Desktop on your company computers, so each person’s AI sessions reach Flowstate. Who: IT and security, with an engineering lead. Where: Roll out Flowstate Desktop Done when: People from your pilot group are counted as active in the opening line on Adoption.

5. Decide who sees names

What: Decide who may see individual people on Adoption: names in the table, who holds idle seats, and each person’s AI record and fluency score. Give those people the View Individual AI Fluency Scores permission. Who: Flowstate admin, with your AI lead and HR. Where: Organisation → People & roles → Roles, or Settings → Users & Access → Roles if your organisation doesn’t use Organisation yet. Done when: Someone with the permission sees names on Adoption, and someone without it doesn’t.

Adoption is for helping teams get going, not for ranking people. The page doesn’t rank anyone by spend, because the heaviest users are usually the ones getting most out of AI.

6. Read Adoption

What: Check the figures look like your organisation, and agree what you’ll act on first: teams that are behind, or seats nobody uses. Who: AI or engineering lead, with your finance lead for seats. Where: Insights → Agent insights → Adoption. See See who uses AI. Done when: How deeply it is used shows people across the levels, and Licensed but idle lists only seats you recognise as unused.

You’re set up when

  • The opening line on Adoption counts active people from every team that has AI seats.
  • Licensed but idle shows seats nobody has opened, not everyone your providers bill for.
  • The people who should see names can, and nobody else.

Reading adoption well

If something’s not right

Nobody shows as active. Sessions aren’t reaching Flowstate. Check Flowstate Desktop is on people’s computers, and that their work email matches. See Roll out Flowstate Desktop.

Everyone your providers bill for shows under Licensed but idle. Same cause: without sessions, Flowstate can’t see that anyone opened their seat.

A team is missing from the table. Its people aren’t allocated to it. See Load your people and teams.

Take-up looks high for a team that relies on contractors. Adoption counts employees only. Group by Employment type to see contractors too; those totals won’t match the rest of the page.

Next