Pocket Resume
Context-aware browser extension that rewrites a master resume to match any job description.
Context & Overview
The modern job market expects a tailored resume for every application to clear ATS filters. Producing those variations manually is slow, repetitive, and easy to do badly.
The Challenge
Problem Statement
Job seekers spend hours rewriting a resume for each application, guessing at the keywords individual ATS systems prioritize.
Why It Matters
Resume tailoring is the highest-leverage step in the application pipeline — the difference between a generic submission and a targeted one directly affects screening outcomes.
Constraints
- Must operate on any job board (LinkedIn, Indeed, etc.)
- Privacy-first: all data stays local to the browser
Goals & Metrics
Objectives
- Reduce resume tailoring from 30 minutes to 30 seconds.
- Generate high-quality, ATS-friendly content.
- Integrate seamlessly into the browsing workflow.
Target KPIs
- Active Users
- Resumes Generated
- Time Saved per Application
Research & Discovery
Methodology
Key Insights
- Formatting matters as much as keywords in ATS screening.
- Users are reluctant to store personal employment data in cloud services.
Approach & Strategy
In-context augmentation: the tool operates directly on the job page, reading the DOM to understand requirements and generating the tailored PDF locally — no copy-pasting, no data leaving the device.
Frameworks
Chrome Extension Manifest V3, Multi-provider AI (Gemini, OpenAI, Anthropic, OpenRouter, custom/local endpoints), Clerk (auth) + Convex (cloud sync)
Collaboration
Open source community feedback on GitHub.
The Solution
A Chrome extension that reads the active tab's job description, takes the user's master resume, and rewrites it to match the role's keywords and tone.
App Visuals
Design Rationale
A browser extension was the only form factor that granted seamless access to job board content without copy-pasting — and local processing satisfied the privacy constraint outright.
Execution
Roadmap
Week 1
Manifest V3 setup & DOM scraping
Week 2
AI prompt tuning
Week 3
PDF generation
Week 4
Store listing
Challenges Overcome
- ⚠ Scraping dynamic SPAs (LinkedIn/Indeed) required robust DOM observers.
- ⚠ In-browser PDF generation needed careful handling of fonts and layout.
Outcomes & Impact
Quantifiable Results
- 1.Live on Chrome Web Store
- 2.Generates a tailored resume in under 20 seconds
- 3.Supports all major job boards
Qualitative Feedback
“Consistently cited the cover letter feature as unexpected value.”
“Strong accuracy in keyword matching reported by users.”
Learnings & Future
Key Takeaways
- Prompt engineering functions as a product feature in its own right.
- Client-side AI is viable and cost-effective at consumer scale.
Next Steps
- Application automation
- Resume score analysis
- History tracking