How It Works
From Sizing to a Private AI Server in Your Office, in Five Steps
A defined process built for regulated firms: we size the system to your data, build and test it before it reaches you, install it with minimal disruption, and manage it every month after.
Start Your Compliance & Sizing ReviewStep 1
Compliance & Sizing Review
Answer three questions: your industry, how many people will use the system, and where your office is. Our sizing page matches your firm to its package on the spot, Sovereign Essentials (up to 10 users) or Sovereign Enterprise (11 to 30), with the price shown up front.
Step 2
Provisioning & Payment
Enterprise-grade GPUs are in high demand. Your hardware & provisioning fee and first month are paid up front through our secure Stripe checkout, by bank transfer (ACH) or card, and your dedicated server is ordered and assigned to your firm as soon as payment clears.
Step 3
Off-Site Build & Testing
We do not assemble anything in your office. Your hardware comes to us first: we stress-test the components, load the AI model sized for your firm (up to the 70-billion-parameter class on larger systems), and configure your private document search before the server ever reaches you.
Step 4
White-Glove Deployment
We deliver and install the pre-configured server in your IT closet, or ship a plug-and-play unit for remote offices. We connect it to your local network, set up your secure document folders, and train your staff. Installation is planned for a single visit with minimal disruption to your day.
Step 5
Managed Sovereignty Plan
AI systems need ongoing care. Under your monthly managed plan we monitor system health, maintain your document index, update your AI models, and keep performance where it should be. Our monitoring sees only whether the system is healthy, never your documents or your staff’s questions.
Every Step on the Record
Project timeline
From sizing to deployment, every step in plain language, with dates.
Uploads and messages
Send us documents and questions in one secure place instead of scattered email threads.
On your phone
The portal works on any device, and you can add it to your phone’s home screen like an app.
Where your system runs, and where your data lives
- On-premise (standard). A dedicated server in your office. Your documents and your staff’s questions are processed on that machine and stay in your building. We monitor it remotely through a secure private connection that reports only system health.
- Fully air-gapped (on request). The server has no internet connection at all. Updates and support happen in person.
Either way, the system runs open AI models on your hardware, so you are not tied to any cloud AI company, and we swap in better models as they arrive.
What we will not do
- Give legal, tax, or compliance advice. We build and manage the technology; your attorney, compliance officer, or accountant decides what your firm’s rules require.
- Promise specific results. Outcomes depend on your firm, your documents, and how the system is used.
- Handle patient or other regulated records without agreed safeguards. Tell us first.