What is TrustLevel?

Making trust measurable and verifiable in a world increasingly shaped by AI.

TrustLevel builds tools and infrastructure that make trust measurable and verifiable. In a world increasingly shaped by AI — where information is abundant but reliability is scarce — knowing what and whom to trust has become the hard part.

Our approach

We don't ask whether something is true. We ask how much it can be trusted — and we treat that as a probability, not a yes-or-no verdict. Any claim, review, or AI output can be examined along four dimensions:

• How confident is it? — Trust is a probability, never a binary.

• What is it based on? — We trace the evidence and reasoning behind a claim.

• Who disagrees? — We surface where sources contradict one another.

• How fragile is it? — We test how conclusions hold when assumptions change.

Where the method comes from

This isn't theory. It grew out of years of hands-on work in Web3 grant programs — evaluating hundreds of funding proposals in Project Catalyst, SingularityNET's Deep Funding, and beyond.

That work taught us a practical lesson: the quality of an evaluation depends on who is doing it and how much their judgment has proven reliable over time. In other words, trust can be modelled as Reputation × Expertise — measured through structured peer review rather than assumed.

From method to product

That principle is exactly what REX puts to work: a peer-review system that scores funding proposals by combining reviewer expertise with earned reputation. It's the most direct expression of our methodology — see REX.

Alongside it, the ZK Voting App applies the same commitment to verifiable trust in a different domain: anonymous, tamper-proof voting on Cardano — see ZK Voting App.

The thesis is simple: trust should be measurable, not assumed. Every product puts that principle to work in a different domain.