ConFit v3: Improving Resume-Job Matching with LLM-based Re-Ranking

2026-05-10

AI Summary

This work addresses the limitations of existing resume-job matching systems, which often lack controllability and interpretability, and highlights that current large language model (LLM)-based reranking approaches inadequately handle noise inherent in real-world hiring data. To bridge this gap, the study presents the first systematic optimization of LLM rerankers for practical recruitment scenarios, introducing an integrated strategy that combines multi-round reranking, listwise reinforcement learning, noise-aware data cleaning, and strong-model distillation. Leveraging Qwen3-8B and Qwen3-32B, the proposed method employs joint supervised fine-tuning and distillation, achieving significant performance gains over state-of-the-art systems and strong baselines—including GPT-5 and Claude Opus-4.5—on a real-world person-job matching dataset, thereby enhancing both matching accuracy and practical deployability.

Abstract

A reliable resume-job matching system helps a company find suitable candidates from a pool of resumes and helps a job seeker find relevant jobs from a list of job posts. While recent advances in embedding-based methods such as ConFit and ConFit v2 can efficiently retrieve candidates at scale, the lack of controllability and explainability limits their real-world adaptations. LLM-based re-rankers can address these limitations through reasoning, but existing training recipes are developed on short-document benchmarks and do not account for noise in real-world recruiting data. In this work, we first conduct a systematic analysis over the LLM re-ranker training pipeline for person-job fit, covering inference algorithm design, RL algorithm selection, data processing, and SFT distillation. We find that using multi-pass re-ranking, training with listwise RL objectives, removing noisy samples, and distilling from a stronger LLM before RL significantly improves re-ranking performance. We then aggregate these findings to train ConFit v3 with Qwen3-8B and Qwen3-32B on real-world person-job fit datasets, and find significant improvements over existing best person-job fit systems as well as strong LLMs such as GPT-5 and Claude Opus-4.5. We hope our findings provide useful insights for future research on adapting LLM-based re-rankers to person-job fit systems.

Problem

• resume-job matching
• LLM-based re-ranking
• controllability
• explainability
• noisy data

Innovation

• LLM-based re-ranking
• person-job fit
• listwise reinforcement learning
• data denoising
• SFT distillation