LLM Resume Analyzer

Role: AI Engineer | Stack: NLP, Semantic Matching, Prompt Evaluation

Developed a resume intelligence tool that scores candidate fit against role requirements and explains the score in plain language. The objective was faster triage for recruiters and clearer improvement signals for applicants.

Scope

Built a hybrid pipeline combining deterministic parsing, vector similarity, and LLM-generated commentary. The system evaluates must-have skills, seniority evidence, project relevance, and keyword quality in one report.

Architecture Snapshot

LLM resume analyzer workflow from resume and job description parsers to vector match, rule engine and feedback output. — Hybrid scoring stack: parser + rules + embeddings + LLM explanation layer.

Impact Metrics

-52% Screening Time

+28% Shortlist Precision

4.6/5 Recruiter Satisfaction

Execution Notes

Added confidence scoring and deterministic penalties for missing critical requirements to reduce hallucinated recommendations. Prompt templates were tuned for concise, role-specific feedback that remains readable by non-technical users.