Get up and running with a private, local AI model on your own hardware. The entry point for individual practitioners, small teams, or organizations exploring private AI for the first time — without sending a single token to the cloud.
All tiers include a final handoff meeting with documentation review and a 30-day warranty period during which minor configuration issues are resolved at no charge.
Lawyers, consultants, researchers, and analysts who need a private AI assistant for sensitive client work — without submitting confidential documents to third-party servers.
Teams currently using ChatGPT, Copilot, or Claude for internal workflows who need to bring that capability fully on-premises due to data sensitivity or compliance requirements.
Organizations that want to validate private AI before committing to a larger rollout. Foundations gives you a working system to evaluate without over-investing in infrastructure.
Healthcare teams, financial advisors, and regulated service providers who need AI capabilities but cannot route patient, client, or transaction data through cloud providers.
Teams that need to query internal policies, contracts, SOPs, or institutional knowledge quickly — grounded in your own documents, not the public internet.
A first-generation AI assistant for a specific department — legal, HR, finance, or operations — before expanding to an organization-wide deployment.
Discovery call, compliance review, hardware specification sign-off. Hardware procurement order placed if needed. Network requirements documented.
LLM server software installed and configured on your hardware. Network access rules applied. Security review completed. Base system tested.
Model deployed and optimized for your hardware. User accounts provisioned. API keys issued. MCP server integration connected and tested.
Onboarding session with your team. Documentation package delivered. System confirmed live. 30-day warranty period begins. 2 included support hours activated.
| Capability | Tier 1 — Foundations | Tier 2 — Professional | Tier 3 — Business | Tier 4 — Enterprise |
|---|---|---|---|---|
| Setup Fee | $8,000 | $18,000 | $38,000 | $75,000+ |
| Deployment Option | Local only | Local or Cloud | Local + Hybrid | HA Multi-Server |
| Models Included | 1 | Up to 3 | Up to 5 | Unlimited |
| User Accounts | Up to 5 | Up to 20 | Unlimited | Unlimited |
| MCP Integrations | 1 | Up to 3 | Up to 5 | Up to 10 |
| Custom Agents | — | 1 | Up to 3 | Up to 8 |
| RAG Pipeline | — | — | Included | Advanced + Hybrid |
| SSO / LDAP | — | — | SSO / SAML | SSO + LDAP / AD |
| RBAC & Audit Logging | — | Basic | Full RBAC + Logs | Full + Compliance |
| Staff Training | — | Admin walkthrough | Half-day (10 people) | Full-day (unlimited) |
| Post-Deploy Support | 2 hours | 4 hours | 8 hours | 30-day hypercare |
Foundations is intentionally lean. These three add-ons deliver the most ROI for teams just getting started with private AI. Multi-add-on bundles of 3 or more are eligible for a 10% discount.
The $8,000 setup fee covers professional services only — labor, configuration, documentation, and post-deploy support. It does not include hardware procurement costs for the on-premises deployment or any cloud GPU instance fees.
Hardware can be provided by your organization (existing hardware is fine for Tier 1), sourced by Creeksea at cost, or financed over 24–36 months. See our Hardware page for current configuration and cost guidance.
All prices are starting rates in USD. Final pricing is determined after a scoping session and provided via a formal Statement of Work. Rates quoted in a signed SOW are locked for the duration of the engagement.
All engagements begin with a complimentary 30-minute discovery call — no pressure, no commitment. We'll discuss your use case, existing hardware, and whether Foundations is the right fit.