RevAIsor awarded Innovate UK grant to solve AI’s “verification crisis” with SENTINEL
A neuro-symbolic architecture built to prove — not just claim — that autonomous AI agents are safe. Inside the Sovereign AI grant funding Project SENTINEL.
London, UK — RevAIsor has been awarded a prestigious “Sovereign AI” grant from Innovate UK to develop Project SENTINEL: a neuro-symbolic architecture designed to deliver provable safety and security for the next generation of autonomous AI agents.
The verification crisis
As AI systems evolve from passive tools into autonomous, goal-directed agents, industry leaders are running into a hard wall — a “verification crisis.” Today’s deep-learning models operate as black boxes, making it nearly impossible to formally guarantee their behaviour or defend against novel threats like prompt injection and data exfiltration.
SENTINEL takes a different route. It pairs a neural “System 1” for fast, intuitive perception with a symbolic “System 2” for deliberative, logical reasoning — a guardrail that can be inspected, audited and proven.
“The UK’s sovereignty in AI depends on our ability to build systems that aren’t just powerful, but provably trustworthy. With the support of Innovate UK, we are building a ‘Symbolic Safety Shield’ that acts as a deterministic guardrail, ensuring autonomous agents always operate within human-defined ethical and security boundaries.”
Dr. Edgar Lopez-Rojas — Founder & Technical Lead, RevAIsor
The team behind the shield
The project brings together a multidisciplinary team redefining AI assurance:
- Dr. Edgar Lopez-Rojas leads the technical vision, bridging complex neural systems and formal verification.
- Matt Decal and Karol See drive the responsible-AI engineering, developing the “Neural Core” and the integration interfaces required for high-performance agentic behaviour.
- Ismaeel Ali leads security engineering, safeguarding the integrity of the symbolic shield against adversarial attacks.
Proof in three months
The three-month Proof of Concept will demonstrate SENTINEL in a simulated multi-agent logistics environment. Crucially, the team will benchmark this deterministic approach against current industry standards like “LLM-as-a-Judge” — showing that symbolic logic offers a superior, auditable and verifiable alternative for high-stakes industries.
“Public funding is a catalyst for deep-tech innovation. This grant allows RevAIsor to accelerate a ‘safe-by-design’ philosophy that will become the cornerstone of our global Agentic Assurance platform.”
The RevAIsor Team
Join the SENTINEL Pilot Program — call for partners
We’re selecting a limited number of partners for the testing and validation phase. If your organisation is developing agentic AI or high-stakes autonomous models and wants to pioneer “safe-by-design” verification, we’d like to collaborate. Selection is ongoing — there’s no fixed deadline.
About RevAIsor
RevAIsor is a UK-based AI assurance startup specialising in governance, risk management, and trustworthy AI architectures. By combining deep technical research with a governance-first mindset, RevAIsor helps organisations deploy autonomous systems with confidence.
About Innovate UK
Innovate UK, part of UK Research and Innovation, is the UK’s innovation agency — inspiring, involving and investing in businesses developing life-changing innovations.
Press / media: contact@revaisor.com