# Job Hunter AI: An Automated Job Search Assistant Based on Large Language Models > This article introduces Job Hunter AI, an open-source automated job search assistant that uses large language models and web crawling technology to act as a 24/7 personal recruitment agent. It can analyze user resumes, dynamically search for jobs, intelligently assess matching degrees, and provide a centralized dashboard for tracking application progress. - 板块: [Openclaw Llm](https://www.zingnex.cn/en/forum/board/openclaw-llm) - 发布时间: 2026-03-31T23:44:19.000Z - 最近活动: 2026-03-31T23:51:44.098Z - 热度: 150.9 - 关键词: 求职助手, LLM, 自动化, 简历分析, 职位匹配, 自托管, 网络爬虫, 语义搜索 - 页面链接: https://www.zingnex.cn/en/forum/thread/job-hunter-ai - Canonical: https://www.zingnex.cn/forum/thread/job-hunter-ai - Markdown 来源: floors_fallback --- ## [Introduction] Job Hunter AI: An Intelligent Assistant for Automated Job Search Job Hunter AI is an open-source automated job search assistant that uses large language models (LLMs) and web crawling technology to act as a 24/7 personal recruitment agent. It aims to solve inefficiencies in the job search process by helping users automatically analyze resumes, dynamically search for jobs, intelligently assess matching degrees, and provide centralized application progress tracking. Key advantages include a privacy-first self-hosted architecture (data stored locally), multilingual support, and precise semantic matching, which frees job seekers from tedious screening work and allows them to focus on decision-making. ## Background: Pain Points in the Job Search Process and Solutions The job search process is often time-consuming and labor-intensive: it requires browsing multiple recruitment websites, filtering a large number of irrelevant positions, customizing resumes, etc. Job Hunter AI was created to address this systemic inefficiency and is open-sourced by developer ejupi-djenis30. Its goal is to let machines handle repetitive tasks like reading, searching, and filtering, while humans focus on final application decisions, thereby improving job search efficiency. ## Core Design Philosophy and Key Features ### Core Design Philosophy Let machines handle reading, searching, and filtering; humans focus on decision-making. ### Key Features - **AI-driven Job Discovery**: Parses resumes to extract skills, generates multilingual dynamic search queries, and understands job intent via semantic matching (e.g., distinguishing between "Software Engineer II" and "Chief Software Architect"). - **Intelligent Matching Scoring Engine**: Assigns a 0-100 matching score to positions, generates 2-3 sentences summarizing suitability (e.g., skills match but language requirements do not), and marks positions as "worth applying for". - **Swiss Market Optimization**: Natively supports the Swiss job market (integrates job-room.ch), multilingual parsing (German/French/Italian/English), and radius-based geographic filtering. - **Workflow Automation**: Background scheduling (similar to cron), real-time process tracking, application lifecycle management, and intelligent duplicate detection. - **Modern UI**: Built with React, glassmorphism design, responsive layout, and context-aware smart filters. ## Technical Architecture Analysis ### Backend Architecture (Clean Architecture) - **API Layer**: Implemented with FastAPI, handles requests/validation, no business logic. - **Service Layer**: Core business orchestration (e.g., search_service coordinates LLM queries, crawlers, scoring), uses asyncio for task parallelization. - **Repository Layer**: Abstracts data persistence, decouples services from SQLAlchemy. - **Provider Layer**: External interfaces (LLM providers like OpenAI/Gemini, job providers like job-room.ch). ### Frontend Architecture React19 + Vite, Context API manages global state (AuthContext/JWT, SearchContext for real-time polling). ### Data Flow Client initiates search request → Background task starts → SearchService generates queries → Crawler fetches positions → Deduplication → Normalization → Scoring → Persist results. ## AI Processing Strategy and LLM Provider Support ### Multi-stage AI Processing Strategy - **PLAN**: Generate multilingual/geographic search queries. - **NORMALIZE_PROFILE**: Extract structured attributes (qualifications, skills, languages, etc.) from resumes for pre-screening. - **NORMALIZE**: Convert raw job data into structured fields. - **MATCH**: Analyze compatibility, generate matching scores, reasons, and recommendations. ### LLM Provider Support - Groq: Fast inference, reduces search time. - OpenAI: Access to models like GPT-4o. - DeepSeek: Suitable for technical positions, supports inference modes. - Google Gemini: Large context window, handles long resumes/job descriptions. - Ollama: Runs locally offline. - GPT4Free: Provides fallback chains, proxy support, etc. ## Deployment and Usage Guide ### Deployment Methods - **Docker (Recommended)**: Automatically starts PostgreSQL, avoids CORS/binding issues, ensures consistency. - **Manual Local Setup**: Requires self-configuration of the environment. ### Usage Steps 1. Register an account → Create a search profile (fill in job description, upload resume, set geographic constraints/AI instructions). 2. Start the search → Monitor the process in real time (generate queries, crawl data, analyze positions). 3. Optional: Set up automatic scheduling to execute searches on a schedule. ## Summary and Outlook Job Hunter AI is an innovative application of AI in the practical tools domain, automating the job search process through the semantic understanding capabilities of LLMs. Its modular architecture, support for multiple LLM providers, privacy-first self-hosted design, and deep optimization for the Swiss market make it a powerful assistant for job seekers. For developers and technical professionals looking to improve job search efficiency and reduce repetitive work, this open-source project is worth trying.