AI Resume Screening System Needed

I need an AI-powered workflow that automatically reviews incoming resumes for my entry-level and mid-level openings, highlights the most relevant candidates, and flags any obvious mismatches. The model should rank applicants primarily on their education background and work experience; other factors can remain optional for now.

Here’s what I expect:
• A lightweight, reproducible pipeline (Python or similar, leveraging NLP libraries such as spaCy, Hugging Face, or OpenAI embeddings) that ingests PDFs, DOCXs, or text and converts them into structured data.
• Scoring logic that weights degree relevance, graduation date, institution, total years in role, and direct title/industry alignment, then returns an ordered shortlist with confidence scores.
• A simple dashboard or JSON/CSV output I can plug into our ATS.
• Clear instructions so my in-house team can retrain or adjust the weighting when hiring goals shift.

Acceptance criteria: the tool must process at least 500 resumes in under an hour on standard cloud hardware, surface the top 10% most qualified applicants, and achieve at least 85% precision compared with a human recruiter’s benchmark set that I will provide.

If you have prior experience deploying similar ML models for resume screening and can showcase a quick prototype, I’d love to see it.