FAQ
What are you building?
An on‑prem, air‑gapped AI platform for IP-heavy SMEs and regulated teams that cannot send sensitive data to cloud AI APIs. Alpha 1 focuses on local inference, model lifecycle/versioning, RBAC + audit logs, and offline updates/diagnostics.
Why will you win against hyperscalers and “sovereign cloud”?
By serving the segment where cloud is not acceptable and proving operational excellence in constrained environments (air‑gapped ops, offline updates/diagnostics) with measurable performance claims anchored to a transparent eval harness.
How do you prevent data leakage?
Alpha 1 assumes no outbound internet. Data stays on customer-controlled infrastructure. Logging and diagnostics are designed to avoid persisting sensitive prompts/responses by default.
Do you depend on proprietary models?
No. The platform standardizes on an open-runtime baseline (Ollama) and supports multiple open model
families. OpenAI (gpt-4o-mini) is used only as a benchmarking baseline for text/domain evals.
What is the business model?
- Software subscription (by node/GPU and feature modules)
- Consulting/services for deployment and operational hardening
- Optional hardware margins via partner bundles/resale
What is the fundraising plan?
Pre-seed equity round. Lean raise: $250k–$300k, framed as ~12 months to hit milestones. Ask size may change with hiring and infra plans.
What are the next milestones after Alpha 1?
- Benchmark report v1 (reproducible, sanitized)
- 2 design-partner pilots
- First paid deployment (subscription + services), even if small