About
We exist to ensure AI is worthy of trust.
With confidence, and with proof.
Making AI deployment decisions defensible.
AI systems are reaching production without structured evidence of how they will actually perform. Benchmark scores are not operational proof. Vendor assurances are not independent verification.
The gap between what an AI system can do and what an organisation can prove it will do is the problem we exist to close.
Built with the backing of those who understand the stakes.
Resaro was established with the support of Temasek, a global investment company headquartered in Singapore.
We are a Premier Member of the AI Verify Foundation, a partner of the Partnership on AI and TUM Venture Labs (Defence/Aerospace).
We help organisations around the world deploy AI they can prove is safe, reliable and fit for purpose.
The people behind the mission.
Making AI deployment decisions defensible requires expertise across engineering, governance, regulation and enterprise transformation. Our leadership team and advisors bring these perspectives together to help shape the future of independent and scalable AI evaluation.
Three principles that shape how we work.
Independent evidence over vendor assurance
The value of AI assurance depends entirely on its independence. Resaro operates as a third party — we have no interest in the outcome of an evaluation other than its accuracy.
In-environment by default
Data residency is not a compliance preference — for the organisations we work with, it is a hard operational requirement. AIP runs inside your infrastructure from day one.
Operational proof over benchmark scores
A model that passes a curated benchmark is not the same as a model proven to perform in the field. Every evaluation we run is designed around the conditions the system will actually face.
Shaping the evaluation standards the industry will run on.
Resaro is actively involved in shaping global AI standards, including an ETSI technical specification that would codify structured evaluation methodology for AI systems, an ISO standard for testing prompt-based text-to-text AI systems, and leadership in the ISO/CEN-CENELEC harmonised standards development process for AI conformity assessment under the EU AI Act.
Our standards engagement serves two purposes: it gives our customers procurement credibility and forward regulatory readiness, and it keeps AIP's evaluation schema aligned to what conformity assessment will require before final standards are published.
We're building the infrastructure layer AI deployment was missing.
We're a growing team working on a consequential problem — in environments where the stakes of getting it wrong are real. If that appeals to you, we'd like to hear from you.