TL;DR: I got tired of spending 2+ hours customizing each job application. So I built an open-source system that lets AI do 90% of the work. Paste a job description, get a tailored resume, cover letter, and interview prep in minutes.
The Problem Every Engineer Knows Too Well
It's 11 PM. You've found your dream job posting. The deadline is tomorrow.
You open your resume. You stare at it. You think: "How do I make my payment system experience sound relevant to their recommendation engine team?"
Three hours later, you've:
- Rewritten 6 bullet points
- Googled "action verbs for resumes" twice
- Questioned every career choice you've ever made
- Still haven't started the cover letter
Sound familiar?
I was doing this dance 3-4 times a week during my last job search. Each application took 2-3 hours of mentally exhausting work. And the worst part? I wasn't even sure if my "customizations" were actually improving my chances.
So I did what any engineer would do: I automated it.
Introducing Career Architect 🏗️
Career Architect is an open-source, AI-powered job application pipeline. Here's the workflow:
1. Paste job description → 2. AI generates everything → 3. Build PDF
That's it. That's the whole process.
What you get:
- ✅ A tailored resume with metrics that match the job requirements
- ✅ A cover letter that complements (not duplicates) your resume
- ✅ Interview prep with likely questions and model answers
- ✅ ATS keyword scoring to optimize your match rate
- ✅ Professional PDFs ready to submit
Time investment: ~10 minutes of review instead of 2+ hours of writing.
How It Actually Works
The "Experience Lake" Concept
The magic starts with what I call your Experience Lake — a comprehensive markdown file containing every achievement, project, and metric from your career.
Instead of trying to remember that time you "improved API latency by 40%" while staring at a blank resume, you build this lake once. The AI then fishes from it for each application, pulling the most relevant catches.
# Master Experience
## Payment Platform (2022-2024)
- Architected distributed payment system processing 50K daily transactions
- Reduced checkout latency from 800ms to 120ms (85% improvement)
- Led team of 4 engineers, improving sprint velocity by 35%
## Real-time Analytics (2021-2022)
- Built streaming pipeline handling 1M events/minute
- Implemented anomaly detection reducing fraud by 23%
...
The Prompt Orchestra
Career Architect uses a series of specialized AI prompts that work together:
| Prompt | Job |
|---|---|
analyser.md |
Gap analysis — what you have vs. what they want |
tailor_resume.md |
Generates targeted resume from your experience lake |
cover_letter.md |
Writes a compelling narrative |
interview_prep.md |
Predicts questions with model answers |
pdf_generator.md |
Validates and prepares for PDF export |
Each prompt has a specific role, like musicians in an orchestra. The main_orchestrator.md conducts them all.
Vertical Targeting
Not all companies speak the same language. A startup wants to hear about your "zero-to-one wins" and "force multiplier" impact. Big Tech wants "scale," "precision," and "methodical approaches."
Career Architect automatically detects company size and adjusts tone:
| Company Type | Tone | Keywords |
|---|---|---|
| 🚀 Startup (<50) | Bold, high-energy | "Shipped," "Owned," "10x'd" |
| 📈 ScaleUp (50-1000) | Collaborative | "Scaled," "Processes," "Cross-functional" |
| 🏢 Big Tech (1000+) | Methodical, precise | "At scale," "Systems," "Reliability" |
The Technical Stack
Career Architect is surprisingly simple:
- Python — CLI tools and scripts
- Markdown — All documents are plain text
- Pandoc + LaTeX — Beautiful PDF generation
- Your AI assistant — Claude, GPT-4, Copilot, whatever you use
No databases. No cloud services. No subscriptions. Everything runs locally on your machine.
Template Gallery
Four professional LaTeX templates included:
| Template | Best For |
|---|---|
| Default | Modern tech companies (slate blue) |
| Minimal | Traditional industries (black & white) |
| Creative | Startups and design roles (bold colors) |
| Executive | Senior/management positions (navy & gold) |
Real Results
After using this system for 3 months:
- Application time: 2.5 hours → 15 minutes (per application)
- Response rate: Increased ~40% (more tailored = more callbacks)
- Interview prep: Actually prepared instead of winging it
- Mental load: Dramatically reduced
The biggest win? Consistency. Every application is now professionally formatted, keyword-optimized, and genuinely tailored. No more "good enough" submissions at midnight.
The Philosophy: Modern Builder
Career Architect includes an optional "Modern Builder" framework for presenting engineering experience. It's based on five core capabilities:
- Precise Problem Decomposition — Breaking complex problems into solvable units
- Systems Thinking — Understanding interconnections and dependencies
- AI Steering — Effectively directing AI tools for development
- Technical Taste — Making sound architectural decisions
- Ownership of Outcomes — End-to-end responsibility for results
Instead of saying "Implemented payment system," you'd say "Locked architectural intent for payment flows, reducing integration entropy across 4 downstream services."
It's a different way of communicating engineering value. Use it if it resonates, skip it if it doesn't.
Getting Started (5 Minutes)
# Clone the repo
git clone https://github.com/henryohanga/career-architect.git
cd career-architect
# Set up virtual environment
python -m venv .venv
source .venv/bin/activate
# Install dependencies
pip install -r requirements.txt
# Verify setup
make check
Then:
- Add your contact info to
source_materials/identity.json - Paste your existing resumes into
source_materials/resumes/ - Tell your AI assistant: "Read
.prompts/setup.mdand build my experience lake"
For each new application:
- Paste the job description
- Say: "Use the Career Architect pipeline"
- Review the generated materials
- Run
make buildfor PDFs
Why Open Source?
Job hunting is stressful enough. Everyone deserves access to tools that level the playing field.
Career Architect is MIT licensed. Use it, fork it, improve it. If you build something cool on top of it, I'd love to hear about it.
What's Next?
Some ideas I'm exploring:
- LinkedIn integration — Auto-optimize your profile for target roles
- Application tracking — Built-in CRM for your job search
- Interview simulator — Practice with AI before the real thing
- Salary negotiation coach — Data-driven negotiation strategies
But honestly? The core system works. It's been battle-tested through dozens of real applications.
Final Thoughts
The job market is competitive. Companies use AI to screen resumes. It's only fair that candidates use AI to create them.
Career Architect doesn't write fiction about your experience. It presents your actual achievements in the best possible light for each specific opportunity. That's not gaming the system — that's effective communication.
If you're an engineer tired of the resume grind, give it a try. Your 11 PM self will thank you.
🔗 Links:
- GitHub: github.com/henryohanga/career-architect
- License: MIT (free forever)
💬 Questions? Open an issue or reach out. Happy to help.
If this helped you land an interview, let me know. Nothing makes an open-source maintainer happier than hearing their tool actually worked. 🎯
Top comments (0)