Zara

Zara sits on top of a genuinely unusual asset. micro1's public research output includes CortexRetrievalBench, an evaluation of how well AI retrieval systems find and rank authoritative sources for financial research, plus Realm benchmarks for legal reasoning and pathology-report reasoning. Building evaluation environments that hold up under expert scrutiny is hard, and micro1 has been doing it in public. Applying that discipline to candidate assessment — score the demonstrated work, not the keyword density on a resume — is a defensible position, and human-in-the-loop validation on the scoring itself is the right design choice for hiring decisions you have to defend later. What you cannot do is evaluate Zara as a product from the outside. micro1's current site is a data-lab site: the navigation is Intelligence, Realm, Cortex, Robotics, Research, with calls to action for experts looking for paid work and for labs wanting data partnerships. Zara does not appear as its own page with features, tiers or integrations. Some of that is a documentation gap rather than evidence of absence — we could not reach the docs or changelog pages, so we make no claim about API availability or release cadence. But a buyer at the evaluation stage still hits the same wall: there is nothing published to compare against Paradox, HireVue or CodeSignal on price, deployment shape or integration surface. The realistic reading is that Zara is a services-and-platform engagement rather than a self-serve product. If you are already talking to micro1 about Realm environments or Cortex evaluations, the conversation is easy to extend and the evaluation methodology transfers directly. If you are not, you will be onboarding into micro1's world to get a recruiter, and that footprint deserves scrutiny. Either way, run it in parallel with your current funnel before you route real candidates through it.