CareerPilotAgent

Local-first AI job matching and application assistant for international business students.

CareerPilotAgent helps users analyse job descriptions, score candidate–role fit, generate realistic resume suggestions, draft editable cover letters, manage an application tracker, and export structured reports. It is designed as a privacy-first portfolio project and does not scrape job boards, auto-apply, or store private resume data in the public repository.

Current checkpoint: CAREER-005-GITHUB-PUBLIC-SHOWCASE

Feature Overview

• Job Description Analyzer — extracts role, company, category, seniority, responsibilities, skills, and keywords
• Explainable Fit Score — presents seven transparent dimensions, matched skills, gaps, and risk notes
• Resume Keyword Suggestions — separates supported keywords from skills that still need evidence
• Project-Based Resume Bullets — generates realistic, copy-friendly portfolio statements without inventing employment
• Editable Cover Letter Draft — creates a conservative draft with placeholders and TXT export
• Application Tracker — stores manual planning data in a local CSV with required-field validation
• Markdown / TXT / CSV Export — produces portable reports and tracker files
• Demo and Local Private Profile Modes — keeps public-safe demo data separate from optional local profile data
• Local-First Privacy Design — requires no account, API key, or external AI service

Why This Project Exists

CareerPilotAgent demonstrates how an AI-style workflow can solve a practical career problem while keeping decisions explainable and user-controlled. The project combines workflow design, product thinking, Business and Marketing context, privacy-aware automation, local data handling, structured report generation, and a portfolio-ready Streamlit dashboard.

Who It Is For

• International students
• Business and Marketing students
• Junior job seekers and career switchers
• AI automation learners
• Recruiters, classmates, potential clients, and portfolio reviewers

Demo Workflow

1. Select Demo Profile or Local Private Profile.
2. Paste a job description or load a synthetic sample.
3. Analyse the job description.
4. Review the fit score, evidence, and gaps.
5. Review keywords, project bullets, and portfolio suggestions.
6. Edit and verify the cover-letter draft.
7. Save a manual tracker entry.
8. Export the Markdown analysis report.

Screenshots

All screenshots use Demo Profile and synthetic sample data.

Home and Demo Profile

Analysis Workflow

See docs/SCREENSHOTS_GUIDE.md for capture and privacy rules.

Safety Boundary

This project does not:

• Scrape LinkedIn, SEEK, Indeed, or other job boards
• Auto-apply to jobs
• Send emails
• Store real private resume data in the public repository
• Require API keys
• Call external AI APIs in the current MVP

It is an explainable planning assistant, not a hiring predictor, legal adviser, immigration adviser, or automated applicant.

Tech Stack

• Python 3.11+
• Streamlit
• pandas
• pytest
• Local JSON and CSV files

Project Structure

CareerPilotAgent/
|-- app.py
|-- career_pilot/             # parsing, scoring, advice, profiles, reports
|-- data/                     # public-safe samples; private/ is ignored
|-- docs/                     # scope, safety, QA, release guidance
|-- portfolio/                # reviewed showcase assets and notes
|-- tests/                    # automated regression tests
|-- tools/                    # public safety checks
|-- PUBLIC_SHOWCASE_MANIFEST.md
|-- PROJECT_STATUS.md
`-- requirements.txt

How to Run

Set-Location F:\AIProjects\CareerPilotAgent
.\.venv\Scripts\python.exe -m streamlit run app.py

Open http://localhost:8501 if a browser does not open automatically.

How to Test

Set-Location F:\AIProjects\CareerPilotAgent
.\.venv\Scripts\python.exe -m pytest

Public Safety Check

Set-Location F:\AIProjects\CareerPilotAgent
.\.venv\Scripts\python.exe tools\public_safety_check.py

The checker scans public text files, skips private/generated directories, reports documentation-only review warnings, and returns a non-zero status for publish-blocking findings.

Public Showcase vs Private Data

• data/sample_profile.json contains synthetic, public-safe demo data.
• data/sample_jobs.csv contains synthetic job descriptions.
• data/private/user_profile.json is optional and local-only.
• data/private/ is ignored by Git.
• Demo Profile must be used for screenshots and public demonstrations.
• Local private profiles, real resumes, contact information, and application notes must never be published.

Git ignore rules reduce accidental commits but do not make local files a secure credential vault.

Documentation

• Public Showcase Manifest
• Screenshot Guide
• GitHub Release Checklist
• Portfolio Positioning
• Safety and Privacy
• Manual QA Checklist

Current Status

CAREER-001 delivered the local MVP, CAREER-002 polished the interface, CAREER-003 added the privacy-first local profile editor, CAREER-004 prepared the release structure and safety tooling, and CAREER-005 published the reviewed screenshot-integrated showcase on GitHub.

Roadmap

• CAREER-005 — complete: Screenshot integration and GitHub publish
• CAREER-006: Optional OpenAI-enhanced JD analysis
• CAREER-007: Interview question generator
• CAREER-008: Resume version matching
• CAREER-009: Compliant job-board manual import

Any future networked feature must remain optional, explicit, and consistent with the project's user-control and privacy boundaries.

Disclaimer

CareerPilotAgent provides educational and organisational support. Its score is a transparent heuristic, not a prediction of hiring outcomes. Users must verify job requirements, protect personal data, edit all drafts, and take responsibility for every claim and application.

Managed through AgentHubControlCenter

This project is part of my local-first AI Agent portfolio and can be managed through AgentHubControlCenter, the central command center for agent manifests, safe actions, useful signals, workflow simulations, connector readiness, approval gates, and public-safe reporting.

CareerPilotAgent is registered as a career operations module in AgentHubControlCenter.