I’m a senior technologist who spent 20 days building a way to kill the "initial phone screen" human-captcha.
Most AI resume tools are just wrappers. I wanted something that actually built trust, so I built a verification engine. When you train your Representative on your history, you attest to specific skills and dates.
When the AI talks to a recruiter, it uses a custom renderer to "stamp" verified claims with a shield badge in real-time. If the AI says I have 10 years of experience in Postgres, it's not a hallucination; it's a data-backed claim tied to my account attestation.
It also handles "Identity Gating"—recruiters have to verify via LinkedIn before they can grab your PDF resume or personal contact info.
Tech: Next.js 16, Supabase (with PG Vector), and Vercel AI SDK.
I'd love a technical roast of:
The verification UX: is self-attestation with a timestamp enough for you to trust a candidate?
The "Representative" concept vs. a traditional static portfolio.
Security of the LinkedIn gate.
I’ll be checking in all day to answer questions.
mdukefirst|14 days ago
mdukefirst|14 days ago
StevenThompson|14 days ago
I wonder what percentage will just "nope" out of interacting with it?
mdukefirst|14 days ago