Your company’s own AI, for 60% less.

Your employees are already using AI on client work, on personal accounts you don’t control. We give the whole firm its own private workspace. We deploy it, optimise it and run it.

Infrastructure held only for you. Models you choose.

  • AWS
  • Microsoft Azure
  • Google Cloud
  • OpenAI
  • Anthropic

One workspace underneath.
Everything your people do on it.

How your firm’s private AI workspace fits togetherFive things your people do in the workspace (drafting and review, research, summarising, comparing documents, whatever comes next) sit on top of one private AI workspace of your own, built and operated by Runvo, so your firm needs no AI team of its own. It runs on single-tenant infrastructure hosted by Runvo in the region you name, shared with no other firm, or inside your own cloud tenancy if your client contracts require it. Everyone at the firm signs in at ai.yourcompany.com with the account they already have.Whatever comes nextComparing documentsSummarisingResearchDrafting and reviewYour own AI workspaceBuilt by Runvo · Run by Runvo · No AI teamWho signs inEveryone at the firmai.yourcompany.comThe sign-in they haveWhere it runsHosted by Runvo, in your regionShared with no other firmOr your own cloud tenancy

Private by design.

Your instance

01

Your firm gets an instance of its own: your own database, your own storage, your own keys. Single-tenant, in the region you choose, shared with no other firm.

Your data

02

Client work stops going through personal accounts on a public AI service. You decide who may see it, how it is protected and how long it is kept.

Your people

03

Everyone at the firm gets an account, not the few who could justify a licence. Private does not have to mean a pilot for six partners.

Two places we can put it

01

Your instance, hosted by Runvo

The default. Single-tenant, in the region you choose, with nothing for your firm to procure and no hardware to buy.

02

Your instance, in your cloud

The same workspace inside your own AWS, Azure or Google Cloud account. It costs more and takes longer. The answer when your own client contracts require it.

The workspace

See it running.

One private AI workspace for the whole firm, at your own address, built and operated by Runvo. Your people stop pasting client work into accounts you cannot see.

ai.yourcompany.com
New chat
Projects
Suppliers12
Board reporting6
Today
Q3 supplier renewal, Ardenne
Expenses policy: taxi limits
Rewrite the job advert
Summarise the board pack
Yesterday
Headcount plan vs budget
CRM options shortlist
Customer update, outage
Ideas for the newsletter
Message
Q3 supplier renewal, Ardenne LogisticsSuppliersOpen-weight 70B
Ardenne want a 6% uplift at renewal. What did we actually agree on price reviews, and is that within it?
No, 6% is above what the contract allows. Clause 8.2 caps an annual review at CPI, and CPI for the reference month was 3.1%.1
Master Services Agreement, clause 8.2“Charges may be reviewed once in each contract year, by no more than the increase in the Consumer Prices Index over the preceding twelve months.”
Two things worth raising before you reply:
1.They already took a 3.4% review in March, so a second one this contract year isn’t available to them.2
2.Notice of any change is 90 days. Renewal is 12 February, so their window closes on 14 November.3
Sources
1Ardenne MSA 2024cl. 8.2
2Rate card historyMar review
3Ardenne MSA 2024cl. 9.1
Reply to Company AI

What your firm gets

  • Everyone at the firm, at your own address
  • Ask, draft, summarise and compare against your own documents
  • Sign-in through the directory your firm already uses
  • The models in use, named, on a page you can send to a client
  • Admin control over who has access and what may go in

How you see a real one

Runvo has no customers yet. You would be the first. So we put one in front of you on a call, and name the models it is calling and where it is hosted.

Not your job

  • You don’t need to understand how any of it works
  • You don’t need to build a specialist team
  • You don’t need to manage another complicated technical project
Arrange a call

Your firm, roughly how many people, and what your staff use today. That is enough.

The shape of a firm this is built for.

  • 320 employees UK

    Recruitment group

    • CV parsing
    • Client submissions
    • Interview notes
  • 180 employees UK

    Accountancy practice

    • Document extraction
    • Return data
    • Client correspondence
  • 260 employees UK

    Management consultancy

    • Proposals
    • Desk research
    • Deck drafts
  • 1,400 employees UK

    General insurer

    • FNOL intake
    • Claim extraction
    • Coverage Q&A
  • 400 employees Spain

    SaaS company

    • In-product assistant
    • Ticket deflection
    • Lifecycle outreach
  • 140 employees UK

    Digital agency

    • Desk research
    • Deck drafts
    • Campaign variants
  • 450 employees UK and Ireland

    Technology recruiter

    • CV parsing
    • Job adverts
    • RTW documents
  • 600 employees Nordics

    Claims specialist

    • Claim extraction
    • Coverage Q&A
    • Fraud triage

Client work is already going into AI you don’t control.

Nobody decided this. People had work in front of them and the AI was there, so drafts, deal papers and case notes go into personal accounts your firm does not administer.

  • One personPastes a client draft into ChatGPT
  • One teamDoes the same. Nobody writes it down
  • Most of the firmIs on accounts you cannot audit
  • One clientAsks, in writing, what AI you use

What nobody in the firm can answer

  • Which accounts hold client material
  • Whether any of it trained a model
  • What to tell a client who asks

A policy telling people to stop does not survive contact with a deadline. What does is one workspace good enough to use.

See what the workspace does

What Runvo is.

Runvo builds and runs a private AI workspace for your whole firm, at your own address. It is single-tenant and it runs where you choose.

Everyone signs in at ai.yourcompany.com. Runvo chooses, runs and updates the models underneath, and says which ones they are.

What it is
A private AI workspace for your whole firm, at your own address, built and operated by Runvo.
What it replaces
Staff using personal AI accounts on client material, and per-seat licences priced beyond a whole firm.
Where it runs
Single-tenant infrastructure, in the region you choose. Your own instance, database, storage and keys, shared with no other firm.
Who operates it
Runvo. No internal AI team required.
Who it’s for
Professional-services firms of 150-800 people in law, accountancy, engineering and consultancy, whose staff handle confidential client work.
What it costs
60% less per person than a premium per-seat licence. The figure is given on the first call.

One engagement. Four steps.

Scope and price are agreed before anything is built. Most firms are live within weeks of signing; inside your own cloud tenancy it takes longer, because it runs at the pace of your change process.

01 Scope

We agree the headcount, whether it runs hosted by us in the region you choose or inside your own cloud tenancy, and which systems it needs to reach. The price is agreed here, so there is nothing to negotiate after this.

What we need from you

How many people, and the name of whoever signs. The order is written before any work starts.

02 Build

We deploy your instance: your own database, your own storage, your own keys, shared with no other firm. It is wired to your directory, so people sign in with the account they already have.

What we need from you

A sponsor on your side, and whoever administers your identity provider for an afternoon.

03 Go live

The workspace opens at your own address for everyone at the firm on the same day. We publish which models it is calling, and we brief your people on what may and may not go into it.

What we need from you

An internal announcement, and a policy line you are willing to put your name to.

04 Operate

We run it. Monitoring, scaling, patching, security, model updates and ongoing management, so your firm never becomes an AI infrastructure team.

What we need from you

The monthly fee, and someone to tell when it is wrong. One predictable cost, per person, per month.

We build it, then we run it.

You know the price before the first step, and it does not move afterwards.

What people use it for, industry by industry.

Recruitment & staffing

  • CV parsing & ranking
  • Candidate submission notes
  • Job spec & advert drafting
  • Interview & call notes
  • Right-to-work documents

Accountancy & advisory firms

  • Client document extraction
  • Client onboarding packs
  • Tax return data assembly
  • Management-accounts commentary
  • Deadline and query correspondence

Consultancies & agencies

  • Proposal & tender drafting
  • Desk research & market scans
  • Workshop write-ups
  • Report & deck drafts
  • Campaign copy variants

Insurance

  • FNOL intake
  • Claim document extraction
  • Policy & coverage Q&A
  • Renewal & cross-sell drafting
  • Fraud signal triage

SMB-focused SaaS

  • In-product assistant
  • Onboarding & activation
  • Ticket deflection
  • Lifecycle outreach
  • Docs & knowledge Q&A

The questions that come up first

If yours isn’t here, ask it directly.

What exactly do our employees get?

A private AI workspace at your firm’s own address, ai.yourcompany.com, and everyone gets it rather than the few people who could justify a licence. They can ask questions, draft and redraft, research, summarise documents, analyse files, work with spreadsheets and use approved firm knowledge. They sign in with the account they already have, through the directory you already run, so joiners get access and leavers lose it the same way everything else works.

Where does it run, and who can see what our people put into it?

By default Runvo hosts it, on single-tenant infrastructure in the region you choose. Runvo operates that deployment, which makes your firm the controller and Runvo the processor: which people at Runvo may reach your instance, what they may do there and how that access is logged are fixed by the engagement agreement, not by a page on a website. Nothing goes to another customer, and nothing is used to improve a product.

Is our client data separated from other firms’ data?

Yes, and that separation is what the word private is doing. Your deployment is single-tenant: your own instance, your own database, your own storage, your own keys, shared with no other firm. There is no pooled index and no shared document store sitting behind several customers at once. It runs in the region you choose.

Do the models train on what our people put in?

No. Runvo does not train models and does not use your material to train anything. A request sent to a model provider is governed by the terms Runvo holds with that provider, and the providers behind your deployment are named in your agreement. If one of your own clients asks you this in writing, the answer is written down in a form you can hand over.

Which models does it actually use?

We tell you in writing before you sign, and again whenever it changes. The workspace reaches both frontier models and open-weight ones, and your people choose per piece of work rather than getting whichever model their subscription came with. Runvo is deliberately model-agnostic: what we agree with you is the quality bar, not a model name, because models are retired and repriced constantly and keeping up with that is a large part of what you are paying for.

Can we have it inside our own cloud instead?

Yes, and some firms have to, because their own client contracts say so. The same workspace is deployed into your own AWS, Azure or Google Cloud tenancy: you hold the infrastructure and Runvo’s access to it is whatever you grant. It costs more and it takes longer, because it runs at the pace of your access controls and your change process. If your contracts do not require it, the hosted default is faster and cheaper.

What does it cost?

Per person it is 60% less than a premium per-seat licence, which is the whole reason the firm can put everyone on it instead of a shortlist. The number itself follows from headcount, so it is a sentence on a call rather than a table on a page: tell us how many people are in the firm and we will tell you what it costs, and it is agreed in writing before anything is built, and it does not move afterwards.

Do we pay for everyone, or only the people who ask for it?

Per person on the workspace, and we agree that number before we start. You can open it to one department first if you would rather. But the price is set where it is so that you do not have to choose: the client work you cannot currently see is not confined to the people who would have put their hand up for a licence.

Is there a limit on how much our team can use it?

No. The price is flat per person and nothing inside the workspace is metered: no message caps, no per-token charges, no usage tiers. If you later point your own applications at the same infrastructure, that is priced separately, because the volume is then driven by software rather than by staff.

Do you have customers we can speak to?

No. Runvo has no customers yet, no case studies and no references, and we would rather say so now than have you find out later. What we can do is show you a workspace running on a call, name the models it is calling and say where it is hosted. Anything published elsewhere under this name belongs to a different company.

What happens if Runvo goes out of business?

A fair question to put to a company with no customers. Your material stays yours: it sits in an ordinary database and object storage in your own deployment, not in a format only Runvo can read, and it comes out. If that risk is the deciding one, deploying inside your own cloud tenancy removes it, because then the infrastructure is already yours and only the operating contract is ours. Ask us to write the notice period, the export and what happens to the instance into the agreement before you sign.

Do we need an AI team to run it?

No. Runvo builds it, hosts it, secures it, updates the models underneath and runs it. There is nothing for your IT function to maintain beyond the identity provider it already runs, and no part of this turns your firm into an infrastructure team.

How is this different from hosting an open-source model ourselves?

Hosting a model is the small half of the job. The larger half is the workspace around it: access through your directory, the tools people actually use, monitoring, scaling, security, model updates, and staff whose job it is to keep all of that running. Runvo does the whole of it against agreed targets for quality, latency, reliability and availability. If you want to own the infrastructure but not operate it, that is the own-tenancy option.

Do we have to cancel ChatGPT, Copilot and the rest?

No, and this is not an argument about your subscription bill. Most firms keep the licences that are genuinely earning their keep and stop buying more, because the workspace covers everyone at a fraction of the price. What matters is that confidential client work moves onto something the firm controls; where the remaining licences sit after that is your decision.

We don’t use AI yet. Is it too early for us?

No. Your people are already using it, on their own accounts, which is the reason to act rather than a reason to wait. Starting from nothing is simpler than starting from sprawl: there is no internal tooling to unpick, and the whole firm begins on the same workspace on the same day.

Is there anything for our team to learn?

No. People open your firm’s AI and start working; it behaves the way they already expect AI to behave. What is new is the line about what may and may not go into it, and we brief your people on that when it opens.

How quickly can we be up and running?

Most firms are live within weeks of signing, because there is nothing for you to build: we deploy your instance, wire it to your directory, and open it at your own address for everyone on the same day. Inside your own cloud tenancy it takes longer, because it runs at the pace of your change process.

What it costs

60% less per person than a premium per-seat licence.

Everyone at the firm, not the few who could justify a licence. Tell us how many people you have and you get the number on the first call, agreed before anything is built.

One price, agreed before anything is built, and it does not move afterwards.

  • We build it.
  • We host it.
  • We secure it.
  • We run it.
  • We keep it current.

The result

One instance, shared with no other firm. We build it for you.

  • Your own instance
  • Your data
  • Your firm’s address
  • Everyone at the firm
  • One monthly invoice

Tell us who you are.

Three fields, all of them required. It takes about twenty seconds, and we reply with the next question.