Vector Hire

🚀 I'm excited to announce Vector Hire, an AI powered resume screening platform that helps recruiters find the best candidates, faster and more accurately.

Vector Hire scores large pools of candidates efficiently using a two-stage ranking pipeline rather than a single AI score.

Stage 1 – Rules Based Scoring: The system applies a series of signal-weighted heuristics – skills, job title, experience, behavioural signals and location – to narrow the ~100,000 candidates down to the top 1,000.

Stage 2 – Semantic Re-ranking: Candidates are re-ranked by the cosine similarity between the candidate embedding and the embedding of the job description (all-MiniLM-L6-v2) for the shortlisted candidates. The final ranking is a combination of both approaches, with 65% rule-based scoring and 35% semantic similarity, to balance precision and context understanding.

This project pushed me to solve problems around resume parsing, scalable candidate ranking, semantic search, AI integration, backend architecture, and delivering a seamless experience for recruiters.

The thing I’m most proud of is building an end-to-end system that goes beyond keyword matching. Vector Hire utilises deterministic ranking and semantic AI to generate valuable insights about candidates, empowering recruiters to make faster, more informed hiring decisions.

To test the project, simply upload one or more resumes and a job description. The system goes through each candidate, creates a blended match score, extracts relevant skills and experience and produces a ranked shortlist with detailed insights to aid hiring decisions.

I am looking forward to your feedback and suggestions for future improvement!ents! 🚀