Launching Voxcruit: AI-Powered Recruitment Platform
Built a 0→1 AI recruitment platform achieving 94% matching accuracy and 2.5x backend performance improvement
94%
AI Matching Accuracy
at platform launch
2.5x
Backend Performance
improvement post-launch
-20%
Release Cycle Time
via Agile governance
The Challenge
Penta Global needed to build Voxcruit — an AI-powered recruitment platform — from zero. The challenge was defining an entirely new product in a competitive market while simultaneously managing a portfolio of national-scale platforms. There was no prior product, no established user base, and no validated AI pipeline.
Discovery & Strategy
I led the discovery phase — conducting competitive analysis, user interviews with HR teams, and technical scoping with the AI/ML engineers. The key insight: existing tools had high recall but poor precision. We needed to flip this by focusing on matching quality over volume. I wrote the foundational PRD and defined north-star metrics around matching accuracy and time-to-shortlist.
Building 0→1
Working across engineering, data science, and UX, I drove the full product lifecycle from concept through MVP to launch. I implemented sprint frameworks, KPI retrospectives, and a phased rollout plan. The AI pipeline was iteratively refined based on real recruiter feedback, improving matching accuracy from ~70% in early testing to 94% at launch.
Results & Impact
Voxcruit launched with 94% AI matching accuracy — a benchmark that exceeded initial targets. The engineering team achieved a 2.5x backend performance improvement through architectural decisions I advocated for during sprint reviews. The product is now a key commercial offering in Penta Global's AI portfolio.