Grey Fox AI
Private, governed AI that does real work. We design and deploy AI systems that sit inside your security boundary: assistants that know your documents, agents that complete tasks end to end, and local models for data that must never leave the building.
Services
AI readiness & use-case review
Where AI will pay for itself in your business, and where it will not.
Assistant & agent builds
Task-specific agents with tools, guardrails, approvals and an audit trail.
Private & local models
Self-hosted models on your own hardware for sensitive workloads.
Knowledge retrieval
Secure search and question-answering over your own documents, with permissions respected.
AI security testing
Prompt-injection, data-leak and misuse testing of AI features before you ship them.
Governance & policy
Acceptable-use policy, data-handling rules and human-oversight design.
Before you ask
Which models do you work with?
We are model-agnostic: leading hosted models where their terms fit your data, and open-weight models run locally where they do not.
Can an agent act without a human?
Only as far as you decide. Every build defines which actions are automatic, which need approval, and which are never allowed.
Ready when you are
Email us the outcome you need. We will tell you honestly whether we are the right fit.