Assertif.AI
A marketplace for verified AI talent. Candidates were assessed on hands-on case studies instead of multiple-choice tests.
The problem
Most AI verifications don't actually test anything. Multiple choice questions, a few videos, a badge. But the real risk to a business isn't whether someone can use AI - it's whether they can tell when it's wrong.
AI skill is mostly self-reported. People list it on LinkedIn, complete short courses and say they can prompt, automate, analyse or code with AI. The question for anyone hiring is whether that person can use AI well in a business context without creating legal, data or reputational risk.
My own research backed this up. A year of Reddit discussion showed hiring managers couldn't verify candidates, and professionals who knew they were skilled couldn't prove it.
What we built
Assertif.AI was a marketplace where businesses could find AI talent they can actually trust, backed by verifications that hold up. Each verification was a hands-on case study built around a real business scenario. Candidates worked through it using AI tools and were assessed against a rubric covering output verification, scope discipline and disclosure.
Verifications covered practical use cases across marketing, finance, legal, operations, coding, HR, data analytics and customer experience.
The bar is set high on purpose. A verification that everyone passes signals nothing.
My role
I co-founded it with Matthew Harris. I owned product, engineering, the AI evaluation engine and the roadmap, and worked on partnerships and go-to-market. The site went live in May 2026, followed by a founding cohort.
Status
Assertif.AI ran from April to September 2026 and is no longer in active development.