◈ Professional services

Bill the client
for the work.
Not for the cloud.

Consultancies, law firms, accountants and agencies increasingly deliver through software: client portals, document review, research assistants, bespoke tools built for one engagement. Compute runs them for less than the leading cloud on the same machines and models, and itemises every charge by project, so infrastructure can be attributed to the engagement that used it.

Used by independent practices, mid-sized firms and global partnerships.

Per-project
Itemised billing
30%
Under the leading cloud’s model service on tokens
Per-second
Machine billing
Zero
Commitments

What this looks like in practice

◆Client portals and secure document exchange◆Document review and summarisation with AI◆Research and knowledge assistants◆Engagement-specific tools and prototypes◆Time, matter and project systems◆Data rooms and archive storage◆Client-facing reports and dashboards◆Client portals and secure document exchange◆Document review and summarisation with AI◆Research and knowledge assistants◆Engagement-specific tools and prototypes◆Time, matter and project systems◆Data rooms and archive storage◆Client-facing reports and dashboards
What holds true here

Four things you stop arguing about

◆

Costs by engagement

Each engagement can be its own project, with its own usage and its own line on the bill.

◇

Review at reading speed

Summarise and classify documents with AI inference billed per token, at 30% under the leading cloud’s model service for the same model.

◈

Client data kept apart

Projects, databases and storage separated per client as a configured property, not a convention.

○

Tools built for one job

Stand up a bespoke tool for an engagement, hand it over or switch it off, and pay only for the time it ran.

The shift

Technology becomes a billable part of the engagement

Firms often absorb technology costs into overheads because they cannot see which client caused them. When every charge is itemised by project at a published price, technology becomes something you can scope, quote and recover.

The usual arrangement
  • ✕Infrastructure absorbed into firm-wide overhead
  • ✕Document review rationed by token cost
  • ✕Client data separated by naming conventions
  • ✕Engagement tools left running after the engagement
  • ✕Quotes that guess at technology costs
What replaces it
  • ✓Usage attributed to the engagement that incurred it
  • ✓Full review at a known price per document
  • ✓Separation configured per client
  • ✓Tools that stop billing when switched off
  • ✓Quotes built from published prices
What you can count on

Stated plainly, and qualified

30%
Under the leading cloud’s model service
General-model tokens
30%
Under the leading cloud on CPU
Same vCPUs and memory, cheapest the leading cloud match
Itemised
By project
Every charge, at published prices
Configured
Client separation
Held to across the platform
Closer in

Professional Services, specifically

The engagement that shows its own costs

A due-diligence exercise needs a data room, a review assistant and a portal for the client. On Compute that is one project: storage in Cloud Services, review on AI inference, the portal on Panel. At the end you can see precisely what it used, recover it from the client if the engagement allows, and close it.

Portals clients trust

Client-facing portals hosted on Panel, on your own domain, with access granted and recorded per person.

Knowledge, searchable

Build assistants over your own precedents and research, with the data staying in your account.

Prototypes in a day

From repository to working demo in an afternoon, for the pitch or the workshop.

Residency per client

Where each client’s data lives is configured and held to, so it matches what your engagement letter says.

Against the leading cloud

Professional Services on Compute, against the leading cloud

General & code model — input
$0.105 / million tokens
Leading cloud $0.15 / million tokens · The same open-weight model, input tokens
30% less
General & code model — output
$0.455 / million tokens
Leading cloud $0.65 / million tokens · The same open-weight model, output tokens
30% less
4 vCPU / 8 GB server
$0.0941/hr
Leading cloud $0.1344/hr · Its cheapest machine with 4 vCPU and 16 GiB
30% less

Compared with the leading cloud's own published on-demand prices, US East. Every comparison.

Questions we are actually asked

Answers without the hedging

Is client data used to train models?+

Your data stays in your account and is used to serve your requests. If you have specific contractual requirements, send them to us and we will tell you plainly whether we meet them.

Can we meet our professional and regulatory obligations?+

Residency, access records and separation between clients are configured properties of your account. For any specific attestation your regulator or clients require, ask us directly and we will tell you what we hold today.

We are a small firm. Is this overkill?+

No. The free allowances cover a small practice’s portal and database, and everything else is billed by use. The same platform serves a sole practitioner and a global partnership.

How do we attribute costs to clients?+

Create a project per client or engagement. Every charge on the bill is grouped by project and priced at a published rate.

Other industries

Where else this goes

AI Agents

Run agents, the tools they call and the models behind them on one platform, billed by the token and the second, for less than the same work costs on the leading cloud.

Customer Acquisition

Host the landing pages, attribution pipelines and onboarding flows that bring customers in, and spend less on the infrastructure than you would on the leading cloud.

Marketing

Run campaign sites, personalisation, content pipelines and marketing data on one platform that costs less than the leading cloud for the same machines and models.

Lead Generation

Run the forms, enrichment jobs, scoring models and routing that turn interest into leads, for less than the same infrastructure costs on the leading cloud.

Sales

Run CRMs, sales data, call analysis and AI assistants on one platform, priced under the leading cloud for the same machines, databases and models.

Creator Economy

Host creator platforms, media processing, storefronts and AI tools on infrastructure that bills by the second and costs less than the leading cloud for the same machines.

Start where it matters most

Begin with one workload. Keep what you have running. Decide whether to continue once you have seen it carry something real.

No migration required to begin. Nothing to tear down if you stop.