Services
How an engagement runs, and what it costs.
Five phases, one published rate, and no stage you cannot see the shape of before you commit to it.
The five phases
Not sure where you are? Almost every engagement starts with discovery, and discovery is deliberately small — it exists so that the decisions after it are made with real information.
- 01
Discovery
We look at how your organization actually works, where AI would earn its place, what it is worth, and what to leave alone. You get a written assessment whether or not you continue.
Typically 20–40 hours depending on the size of the organization.
- 02
Proposal
The plan, the architecture, and an implementation estimate grounded in what discovery found. This is where a real number appears — before you have committed to the work it describes.
Typically 10–20 hours.
- 03
Implementation
The build. Depending on what the proposal scoped, this can include standing up a private PeerLLM deployment across your own machines, custom applications and integrations, data preparation, or coaching your teams through the change.
Estimated in the proposal, with a checkpoint conversation before any overrun.
- 04
Support
Keeping what we built correct as models improve, requirements shift and people move on. Ongoing, and the reason most of this work holds up a year later.
- 05
Upgrades
New model versions, new platform capability, new opportunities that did not exist when we started. Adopted on your schedule.
Rates
One rate, published.
$300per hour, all phases
Most consultancies will not tell you this until you have sat through a call. We would rather you could work out roughly what an engagement costs before you contact us — which is why the hour ranges above are published alongside the rate.
Founding customers
PeerLLM is new, and we are taking on our first business clients now. Founding customers get 25% off the published rate and direct involvement from the person who built the platform, in exchange for being first — and for letting us describe the work afterwards.
It is a genuinely better deal than we will be able to offer later.
What implementation can include
- Consulting
- Where AI applies, what it is worth, and what to leave alone.
- Deployment
- Standing up a private PeerLLM network across the compute you already own.Covered in detail here.
- Custom development
- Applications, integrations and data work built around the network.
- Coaching and culture
- Working with your teams until AI is part of how they actually operate — including being honest about the tasks it is wrong for.
Two things both called support
Product support for PeerLLM itself is included with a current licence and comes from the platform, not from us as a billed engagement.Ongoing engineering — on the applications, integrations and data work we built around it — is billed here at the rate above.
They get confused constantly in this industry. We would rather draw the line before you sign anything than after.
Where this has been done before
The largest system we have built is one you can go and inspect yourself:PeerLLM, a live decentralized inference network with published health and fairness metrics.