Resaro Approved Intelligence Platform and ASQI

Resaro is an independent, third-party AI assurance provider established by Temasek, with headquarters in Singapore and Germany. It tests and evaluates mission-critical AI systems for defence agencies, government bodies, critical infrastructure operators, healthcare and financial services, and is a premier member of Singapore's AI Verify Foundation.

The Approved Intelligence Platform runs testing, evaluation, validation and verification (TEVV) workflows inside the client's own environment, including air-gapped and on-premise setups. It builds versioned golden datasets from text and image data plus synthetic edge cases, runs repeatable test packages for performance, safety and security, and turns results into auditable deployment evidence and leadership summaries structured by Resaro's AI Solutions Quality Index (ASQI), which is also released as open source. Test packages plug into CI/CD for continuous evaluation after deployment. Clients cited include HTX, DSTA, NHG and Tookitaki.

• Golden datasets: versioned test data from text and images, extended with synthetic edge cases.
• TEVV workflows: define criteria, test packages and thresholds for performance, safety and security.
• Decision support: auditable deployment evidence traced from outputs back to inputs.
• ASQI framework: common quality index for technical, leadership and governance teams, available as open source.
• Secure deployment: runs air-gapped and on-premise in the client's own environment.
• CI/CD integration: test packages plug into automation for continuous post-deployment evaluation.

Vietnamese banks can obtain independent validation of AI-based AML investigation assistants before rollout. Public security and defence agencies can test vision and deepfake detection systems in air-gapped environments. Hospitals and health tech firms can validate clinical AI tools against agreed quality criteria.

Buyers get independent, third-party evidence on whether an AI system is fit for its specific use case before signing off deployment. Testing runs inside the client's environment, so sensitive data and models never leave secure or air-gapped networks. Reports translate technical results into clear operating conditions and residual risk for leadership. Continuous evaluation catches degradation as conditions and threats change.

Singapore's AI Verify Foundation and IMDA are building an ecosystem of independent AI testers, and governments across ASEAN are drafting AI rules that call for evidence-based assurance. Public agencies and regulated firms in Vietnam adopting AI for citizen services, AML and healthcare will need third-party validation before deployment.

Established by Temasek, co-headquartered in Singapore and Germany. Premier member of the AI Verify Foundation and member of KI Park and the Partnership on AI. Part of NVIDIA Inception and linked with TUM Venture Labs. In September 2025 it partnered with Temus to act as independent evaluator of AI solutions.