SkillVee

Work simulations that show who can actually do the job

www.skillvee.com

The Problem

Hiring is broken in the AI era

72%of resumes are now AI-written and look alike
45%of candidates use AI during interviews
<1%of applicants are truly qualified
The key question remains: Can this person actually do the job?

The Solution

A 45-min simulation of the real job

Live AI stakeholders

Voice conversations with PM, manager, and tech lead — with memory and context

Real deliverables, not trivia

Candidates produce actual work output — code, PRs, and decisions

Full capture & structured evaluation

Screen + voice recorded end-to-end, with scorecards, transcripts, and artifacts

Compare candidates side-by-side on real work.

Why Now

Live AI models

Real-time voice + context make realistic stakeholder conversations possible at scale.

Multimodal AI

Analyze hour-long work sessions reliably — see how candidates actually work.

This was impossible 6 months ago. Now it works.

Key Insight

Interviews are going away. The future of hiring is watching people work.

Resumes, take-homes, and interviews are all gameable now. Watching real work isn't.

Market

$3B pre-hire assessment market. Growing 16% CAGR.

Comparables: Mercor ($10B), Juicebox ($850M) — AI solved sourcing. HackerRank ($500M, $221M rev) — the incumbent we replace. Evaluation is the missing layer.

Moat

Every simulation teaches us what "good" looks like for each company. As we track which hires succeed, we can predict who will thrive where — not generically, but for each team specifically.

Competitors can copy the product. They can't copy the data. Starting with engineering, expanding to every white-collar role.

Business Model

SaaS with freemium. Paid plans from $79/month to enterprise custom pricing.

Beachhead: Engineering teams hiring 5+ developers/month — hardest roles to evaluate, highest cost of a bad hire. Expands to design, product, and all white-collar roles.

Progress

Working prototype built. 2 committed pilots. Full simulation flow live — AI voice conversations, real coding environment, screen capture, and automated scorecards.

Stack: Next.js, Supabase, Gemini Live (voice), Gemini Flash (evaluation). Built and shipping fast.

Team

German Reyes

German Reyes CEO

  • 2 HR Tech exits (acq. by Buk)
  • Built to $1.4M ARR
  • ex-GPM, Walmart eCommerce
  • Berkeley MBA
Matias Hoyl

Matias Hoyl CTO

  • Meta GenAI
  • Built Zapien (35K users)
  • Stanford MS
  • 2x Stanford Learning Design winner

We know the buyer

HR Tech exits

We can build this

Meta GenAI

We move fast

Working prototype