Seeking UK Deployer Partner for BridgeAI AI-Agent Assurance Pilot

Exovat Limited is seeking an eligible UK deployer organisation to collaborate on an application to the BridgeAI Deployment Catalyst: AI for Assurance and Compliance competition.

ABOUT THE SOLUTION

Exovat is developing Decirance, an open-source AI-agent assurance framework and working demonstrator. Decirance brings together evaluation results, technical controls, operational evidence and human accountability to determine whether a specific AI agent should receive a context-bound Deployment Permit.

Its Assurance Delta capability assesses changes to an agent’s model, prompts, tools, permissions, integrations or operating environment. It identifies which assurance claims and evidence remain valid, what requires reassessment and whether the existing permit should continue, become restricted or be suspended.

Decirance is also developing AI-assisted capabilities for evidence extraction, evidence-to-claim mapping, contradiction detection and assurance-test generation. These outputs remain subject to human review. Final permit decisions are deterministic, traceable and approved by a named accountable person.

WHO WE ARE SEEKING

The deployer must be either:

• a UK registered medium or large business; or

• an eligible UK-based public-sector organisation.

The organisation should operate in one of the BridgeAI eligible sectors:

• clean energy industries;

• financial services;

• health and life sciences;

• professional and business services;

• advanced manufacturing;

• defence; or

• creative industries.

The ideal partner is developing, piloting or preparing to deploy an AI agent in a regulated, safety-relevant or consequential operational workflow.

THE PROPOSED COLLABORATION

During BridgeAI Phase 1, Exovat and the deployer would jointly:

• define a credible operational use case;

• assess the readiness of the AI solution, supplier and deployer;

• identify technical, regulatory, governance and assurance barriers;

• create an assurance case for the selected AI agent;

• demonstrate Decirance using representative evidence, workflows and systems;

• introduce controlled changes to the agent’s model, prompts, tools or permissions;

• evaluate which evidence can be safely reused and what must be reassessed;

• measure assurance effort, change-detection accuracy and deployment readiness;

• develop a pathway towards live deployment and replication.

WHAT EXOVAT PROVIDES

Exovat will provide:

• the Decirance framework and demonstrator;

• AI-agent assurance architecture;

• technical implementation;

• assurance-case and Deployment Permit methodology;

• AI-assisted evidence-analysis development;

• project and delivery management;

• a UK-based multidisciplinary core team covering AI implementation, product development, backend engineering, data analysis and delivery.

WHAT WE NEED FROM THE DEPLOYER

The deployer should contribute:

• a meaningful AI-agent deployment use case;

• relevant operational and governance stakeholders;

• access to representative workflows, systems and processes;

• representative technical, evaluation and assurance evidence;

• participation in readiness and benefits assessment;

• support for a controlled demonstration;

• contribution to the project scope, costs and BridgeAI application;

• willingness to consider subsequent deployment phases if invited.

Phase 1 does not need to begin with unrestricted production access or personal data. The initial work can use representative, synthetic, redacted or appropriately controlled evidence where necessary.

We welcome discussions with organisations that want to move an AI agent from experimentation towards trusted, accountable and evidence-based operational deployment.