AI Candidate Recommendation Bot

I need an AI-driven bot that takes the heavy lifting out of recruiting by automatically recommending the best-fit candidates for each open role. The bot’s single, measurable outcome is simple: generate a ranked shortlist so I can move straight to interviews instead of wading through piles of résumés.

Here’s the workflow I have in mind. A hiring manager uploads or points the system to application data—CV PDFs, LinkedIn profile links, or any structured form we decide to support. Your model then parses, normalises, and enriches that data, evaluates each applicant against the role description, and returns a clear, ordered list with confidence scores plus a quick justification (skills matched, years of experience, keywords, etc.). Weightings for experience, education, and specific skills must be easy for me to tweak per vacancy, ideally through a small dashboard or config file.

While I’m open to the tech stack, Python with libraries such as spaCy, scikit-learn, or OpenAI embeddings feels natural; if you prefer another modern NLP toolkit, just explain why. The bot will ultimately sit behind my existing ATS, so clean RESTful endpoints and concise documentation are a must.

Deliverables
• End-to-end codebase (data ingestion, ranking engine, API layer) in a private repo
• Minimal UI or Postman collection demonstrating the recommendation call and response
• Setup guide and a short report explaining model choice, feature engineering, and how to adjust weighting parameters
• One feedback round after I test with real candidate data

Acceptance criteria: on a supplied sample of 100 applicants, the top-10 list must capture at least 80 % of the candidates manually flagged as “interview-worthy.”

If this sounds straightforward for you, tell me how you’ll tackle CV parsing reliability, bias mitigation, and integration into an ATS environment, along with an estimated timeline.