The technology behind our products
Our products run on technology we design ourselves: Prime Edge AI for deciding where AI requests are answered, Green AI for using less energy per request, secure infrastructure for AI workloads, governance for AI agents, privacy engineering for sensitive data, and risk tools for health and insurance. We have filed three UK patent applications to protect it.
Last updated
Our technology
-
Prime Edge AI
Decides where each AI request is answered
-
Patent portfolio
Three UK patent applications
-
Secure AI infrastructure
Zero-trust design and control over where AI requests go
-
Anonymisation and privacy engineering
Anonymisation, pseudonymisation and redaction for sensitive data
-
Green AI
Our work on using less energy per AI request
-
Agent governance
Logs, tests and reviews what agents do
-
Risk intelligence
Risk stratification, risk management and claims integrity
Where AI is going, and what we are building for it
Six shifts in AI, and our own work on each one.
-
AI agents
AI that carries out set tasks, not only answers questions.
What we build Agents with a named job, a person approving what matters, and a log of every action.
Our AI agents -
Small models on devices
Capable models now run on phones and laptops.
What we build Prime Edge AI answers each request on the device when it can, and in the cloud only when needed.
Prime Edge AI -
Green AI
The energy each AI request uses is becoming something buyers measure.
What we build Our work on using less energy per request, measured with a published method.
Green AI -
Agents that use your tools
Open standards such as the Model Context Protocol (MCP) connect AI to the systems a business already runs.
What we build Agents connected to Salesforce and your other systems, through their own interfaces and MCP.
Digitisation and Salesforce -
AI assurance
Buyers and regulators want evidence that AI is tested and controlled.
What we build Testing before launch and after every change, audit trails, and clinical safety cases under DCB0129.
Responsible AI -
Voice AI
Talking is becoming a normal way to use software, in healthcare too.
What we build Prime Assist: AI voice infrastructure for healthcare, answering calls in the caller's own language.
Prime Assist
Principles we build by
- Privacy by architecture
- Routine processing stays on the device wherever it can, so there is less personal data to protect in the first place.
- Safety designed in
- Clinical safety is part of the design from the first day, not a review added at the end.
- Secure by default
- Every request is verified, and every component gets only the access it needs.
- Measured, not assumed
- We measure how our systems perform against agreed goals, rather than assuming they work.
- Human oversight where it matters
- People review the decisions that have consequences for people.
Technology briefs
One-page briefs on how our technology works, each with a full HTML version.
- Prime Edge AI technology brief September 2026
Download the brief (PDF, 1 page)