About the Search & Recommendation category

Matching intent with inventory. The systems that connect users to the right content at scale

This is a top-level domain. Subdomains open as new evaluations roll out. Badges are earned at two levels: component and subdomain. Each badge represents a verified reasoning evaluation that can be shared as proof of how you think.

What evaluations look like

  • Reasoning interviews, not coding tests
  • 3–5 main questions answered via written chat, with adaptive follow-ups
  • Short voice reflection at the end
  • 2 attempts per evaluation; last attempt counts
  • Detailed dimension-level feedback after every question. Evaluations double as upskilling
  • Free for candidates

Visibility controlled by candidates

What hiring looks like

See plan details →

  • Search existing evaluations. Apply standard filters, or use Ask AIVIA to search reasoning traces and find engineers whose thinking patterns match what you need.
  • Run your own evaluations. Fork an existing AIVIA evaluation, generate one from scratch, post it publicly, or send evaluation links directly to candidates.
  • Build your shortlist. Save candidates from existing evaluations, posted evaluations, or direct links you sent.
  • Run integrity checks before interviews. Generate interview questions tailored to each candidate’s evaluation, review voice reflections against transcripts, or request a follow-up evaluation when needed.
  • Re-rank with your dimensions. Compare candidates against the technical dimensions that matter for your role.


How AIVIA evaluates →
Launch offers →
Quick Start →
Other FAQ →