Your team's AI assistant, inside your own cloud
A ChatGPT-style assistant that answers from your firm's documents, runs in your Azure, AWS or private server, and never sends client data to a public AI service. Installed, secured and handed over in two to three weeks.
Sound familiar?
- Staff are already pasting client material into public AI tools, and you cannot see what went where.
- Your regulator, insurer or a client questionnaire is asking what your AI policy is.
- The answers your team needs are buried across SharePoint, Drive and email, and search does not find them.
What you get
- Chat interface your staff already know how to use, with single sign-on to Microsoft 365 or Google Workspace
- Model hosted in your own cloud tenant and region, or fully on-premise where required
- One document source connected (SharePoint, OneDrive, Google Drive or a file share) with each user seeing only what they already have access to
- Prompt logging, retention rules and redaction for client identifiers
- A written security summary you can hand to your regulator, insurer or a client
- 30-minute staff training and 30 days of support
Not included
- Additional document sources and custom agents (quoted separately)
- Model fine-tuning
- Any processing of your data on our systems — everything runs in yours
Pricing
- Start here
- AI Readiness Call + written plan · $500
- A 30-minute call and a written plan: which documents, which model, where it runs, what your regulator or insurer will ask, and a fixed quote. Credited in full against the build.
- Fixed-price build
- $8,000–12,000
- Typically 2–3 weeks. Quoted from the audit, not a rate card. 50% to start, 50% at handover.
- Optional monthly management
- $800–1,500 / month
All prices in USD. Paid securely via Razorpay.
How it works
- 01
Readiness call
30 minutes, then a written plan and fixed quote within three working days.
- 02
Deploy
We install and configure in your cloud with a scoped, time-limited admin account you control.
- 03
Secure and connect
SSO, permissions, logging and your first document source.
- 04
Hand over
Training, the security summary, and a runbook. Optional monthly management after that.
Async by default: one 30-minute kickoff, written weekly updates, no standing meetings.
Proof, not promises
We run our own products on the same stack we install for clients, and we manage production servers for research and analytics businesses today.
See the work →Questions
- Does our data leave our environment?
- No. The assistant, the document index and the model endpoint all run in your cloud account or your own server. We never copy client data to our systems.
- Which models do you use?
- Your choice per regulator and budget: a private endpoint of a frontier model inside your own Azure or AWS tenant, or an open model running entirely on your hardware.
- Do we need an IT team?
- No. That is who this is for. The optional monthly plan covers updates, backups, model upgrades and a usage report.
- Who are we contracting with?
- FutureGen Systems, with an engagement letter, a data processing agreement and a written scope before any work starts.
Start with the $500 readiness call
Fixed scope, written plan, credited in full if you go ahead. If the numbers don't work, we'll tell you.