CareerOps AI — Autonomous Job Search Agent

An end-to-end AI-powered job search assistant that scrapes remote job boards, scores each listing against your CV, researches companies, identifies skill gaps, generates cover letters, and presents everything in a premium web dashboard — all with minimal manual effort.

Cost: ~$0.02–0.05 per pipeline run (50–200 jobs scored with GPT-4o-mini). Cover letters use GPT-4o at ~$0.01 each.

---

Screenshots

Dashboard — Job Listings	Market Gap Report

---

Architecture

flowchart LR
    subgraph Input
        CV[PDF CV Upload]
    end

    subgraph Pipeline
        S[7 Job Scrapers]
        P[CV Parser<br/>PyMuPDF + GPT]
        SC[Scoring Agent<br/>GPT-4o-mini]
        R[Research Agent<br/>DDGS + GPT]
    end

    subgraph Storage
        DB[(SQLite DB)]
    end

    subgraph Output
        D[Web Dashboard<br/>Flask + Tailwind]
        G[Gap Report<br/>GPT-4o-mini]
        CL[Cover Letters<br/>GPT-4o]
    end

    CV --> P --> DB
    S --> DB
    DB --> SC --> DB
    DB --> R --> DB
    DB --> D
    DB --> G --> D
    DB --> CL --> D

---

Features

• Multi-source job scraping — pulls listings from 7 job boards (Remotive, Himalayas, Jobicy, Arbeitnow, RemoteOK, WorkingNomads, WeWorkRemotely)
• AI fit scoring — GPT-4o-mini scores each job 0–100 against your parsed CV with matched/missing skills and a rationale
• Company research — DuckDuckGo + company website → GPT summary of description, culture, tech stack, funding, and red flags
• Market gap report — aggregates skill gaps from shortlisted jobs only, includes upskill recommendations with resources
• Cover letter generation — GPT-4o writes tailored cover letters per job
• Application pipeline — track status (Discovered → Applied → Interview → Offer → Rejected) with per-job notes
• Smart CV switching — uploading a new CV wipes old data for a clean, domain-relevant start
• Web dashboard — Flask + Tailwind CSS UI with tabbed layout, filters, status pills, KPI cards, and intel panel

---

Project Structure

.
├── agents/
│   ├── scoring_agent.py        # CV vs JD fit scoring (gpt-4o-mini)
│   ├── research_agent.py       # Company research via DDGS + website (gpt-4o-mini)
│   ├── gap_report_agent.py     # Market-level skill gap analysis (gpt-4o-mini)
│   └── cover_letter_agent.py   # Tailored cover letter writer (gpt-4o)
├── scrapers/
│   ├── base_scraper.py         # Shared scraper base + HTML stripper
│   ├── remotive.py
│   ├── himalayas.py
│   ├── jobicy.py
│   ├── arbeitnow.py
│   ├── remoteok.py
│   ├── workingnomads.py
│   └── weworkremotely.py       # Playwright-based (JS-rendered site)
├── pipeline/
│   └── runner.py               # Orchestrates scrape → persist → score
├── parsers/
│   └── cv_parser.py            # PDF CV parsing via PyMuPDF + GPT
├── database/
│   ├── schema.py               # SQLite schema definitions
│   └── db.py                   # All DB helper functions
├── dashboard/
│   ├── server.py               # Flask API server
│   └── index.html              # Single-page dashboard (Tailwind CSS + JS)
├── config.py                   # Central config (models, thresholds, sources)
├── .env.example                # Environment variable template
└── requirements.txt

---

Quick Start

1. Clone and set up environment

git clone https://github.com/Hira-1/career-ops-ai.git
cd career-ops-ai
python -m venv venv
# Windows
venv\Scripts\activate
# macOS / Linux
source venv/bin/activate
pip install -r requirements.txt

2. Install Playwright browsers

playwright install chromium

3. Configure environment variables

cp .env.example .env

Edit .env and add your OpenAI API key:

OPENAI_API_KEY=sk-...
SHORTLIST_THRESHOLD=65        # jobs at or above this score are shortlisted
COVER_LETTER_THRESHOLD=70     # minimum score to generate a cover letter

4. Upload your CV

Start the dashboard and upload your CV via the UI (PDF only):

python -m dashboard.server

Open http://localhost:5000 → click Upload CV in the header.

5. Run the pipeline

Click Run Pipeline in the dashboard header, or run directly from the CLI:

python -m pipeline.runner

This will:

1. Scrape all configured job boards
2. Score new jobs against your CV
3. Save results to data/careerops.db

6. Explore results

Refresh the dashboard. Use the sidebar filters, click any job to view the full intelligence panel, and switch to the Market Gap Report tab for skill gap analysis.

---

Configuration

All settings live in config.py and can be overridden via .env:

Variable	Default	Description
OPENAI_API_KEY	—	Required. Your OpenAI API key
SHORTLIST_THRESHOLD	65	Minimum fit score to shortlist a job
COVER_LETTER_THRESHOLD	70	Minimum score to auto-generate cover letters

Model routing (edit config.py directly):

Task	Model
Job scoring	gpt-4o-mini
Company research	gpt-4o-mini
Gap report	gpt-4o-mini
Cover letter writing	gpt-4o

---

Dashboard

The web dashboard runs at http://localhost:5000 and provides:

• KPI strip — total jobs, shortlisted count, average fit score, top score
• Sidebar — search, score filter slider, status filters, candidate profile summary
• Ranked Job Listings tab — sorted by fit score with matched/missing skills, salary, location mode badges, AI rationale, status selector, notes
• Selected Job Intelligence panel — detailed score breakdown, AI rationale, company insights (description, culture, funding, tech stack, red flags, website), cover letter generator
• Market Gap Report tab — skill development plan (merged gaps + recommendations with clickable resources), market trends, in-demand skills, your strengths
• Application Pipeline strip — live count per status stage

API Endpoints

Method	Endpoint	Description
GET	/api/data	All dashboard data in one call
POST	/api/status	Update job application status
POST	/api/notes	Save notes for a job
POST	/api/cover_letter	Generate cover letter for a job
POST	/api/pipeline	Trigger scrape + score pipeline
POST	/api/regen_gap	Force-regenerate market gap report
POST	/api/upload_cv	Upload and parse a new CV (PDF)

---

Job Sources

Source	Type	Notes
Remotive	REST API	Remote tech jobs
Himalayas	REST API	Remote jobs with salary data
Jobicy	REST API	Remote jobs
Arbeitnow	REST API	European + remote jobs
RemoteOK	REST API	Remote developer jobs
WorkingNomads	REST API	Remote jobs for nomads
WeWorkRemotely	Playwright	JS-rendered, requires Chromium

---

Running Tests

pytest tests/ -v

---

Notes

• The SQLite database (data/careerops.db) and your .env file are excluded from git via .gitignore
• Company briefs are cached in the DB — the research agent skips companies already researched
• The gap report is cached and only regenerated when explicitly requested via the dashboard or force=True
• Playwright scraping respects polite delays (3–5 seconds between requests)

---

Requirements

• Python 3.11+
• OpenAI API key (GPT-4o-mini + GPT-4o access)
• Chromium (for WeWorkRemotely Playwright scraper)

---

Future Work

• Scheduled pipeline runs — cron/APScheduler to auto-scrape daily
• Email/Slack notifications — alert when new high-score jobs appear
• Multi-user auth — support multiple users with separate CVs and pipelines
• Job board pagination — handle 500+ results per source
• Resume tailoring — auto-adjust CV highlights per application
• Interview prep agent — generate likely interview questions per company/role

---

Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

---

License

This project is licensed under the MIT License — see the LICENSE file for details.