Data Handling and Engagement Terms
You should not have to choose between speed and protecting your IP. Written commitments and enclave architecture: NDA-first, zero training on client data, no commercial LLM exposure.
Speed without careless speed
You should not have to choose between speed and protecting your IP. This page explains how we use AI-assisted research and drafting without exposing your documents, prompts, pricing logic, or capture strategy to commercial AI platforms. Our IP-protection posture is contractual, not aspirational — the commitments below are written into every engagement letter and survive the engagement. The enclave architecture exists specifically so these commitments are technically enforceable, not just legally binding.
Data handling
- NDA-First — Mutual NDA executed before any substantive discussion of capture strategy, technology, or competitive position.
- Zero Commercial LLM Exposure — All AI-accelerated research runs on our private on-premise enclave. Your queries, documents, and deliverables never traverse ChatGPT, Copilot, Gemini, or any commercial LLM API.
- Zero Training on Client Data — Client material is never used to train, fine-tune, or evaluate any model. Engagement data is segregated by client and access-controlled.
- Destruction on Close — On engagement close, working copies of client material are destroyed on a documented schedule. Final deliverables are retained per the engagement letter.
Engagement terms
- Conflict-Free — We don't represent competing bidders on the same program. If a conflict emerges mid-engagement, we disclose and decline.
- Senior-Only Delivery — The principal advisor delivers the work. No analyst pyramid, no offshoring, no handoff.
- Scoped Outcomes — Every engagement names a deliverable and the decision it enables. We don't bill against open-ended hours.