btr-resume 1.0.3

AI-Powered Resume Builder - Polish, tailor, and optimize your resume locally

BT Resume

A free, local AI resume helper that polishes your resume bullets, tailors your resume to job descriptions, and formats new experience — all running privately on your own machine. No API keys, no subscriptions, no data leaving your device.

---

Quick Start

Prerequisites

• Python 3.10+ — Download Python
• Ollama — Download from ollama.com (handled automatically on first run)

Install & Run

pip install btr-resume
btr serve

That's it. Your browser opens to http://localhost:5000 with the BT Resume web UI.

First run: btr serve will automatically attempt to start Ollama and download the Mistral 7B model (~4.1GB). If Ollama is not installed, you'll be prompted to install it from ollama.com.

---

Why BT Resume?

Most AI resume tools cost money, require an account, or send your personal data to external servers. This is my take on a free alternative — built from scratch in Python + React, powered by open-source LLMs that run entirely on your device.

---

Features

• Bullet Polish — rewrites weak resume bullets into strong, ATS-optimized ones using proven resume writing formulas
• Job Tailoring — aligns your resume language and keywords to match a specific job description without changing your real experience
• Experience Updater — describe a new project or role in plain English and get polished, formatted resume bullets back
• Resume Grading — get scores and feedback on your resume's effectiveness
• PDF + TXT input — works with both plain text and real PDF resumes
• PDF output — generates a clean, formatted PDF resume ready to send
• 100% offline — no internet connection required after setup, no API keys, completely free

---

Commands

Command	Description
btr serve	Start the server and open the browser (same as plain btr)
btr	Start the server and open the browser
btr setup	Manual setup: install Ollama and download the AI model

---

Configuration

BT Resume is zero-config by default, but you can customize via environment variables:

Variable	Default	Description
BT_RESUME_HOST	127.0.0.1	Server bind address
BT_RESUME_PORT	5000	Server port
BT_RESUME_DEBUG	false	Enable debug logging
BT_RESUME_OLLAMA_HOST	http://127.0.0.1:11434	Ollama server URL
BT_RESUME_ENABLE_ANALYTICS	true	Anonymous usage tracking (set to false to disable)
BT_RESUME_SECRET_KEY	(auto-generated)	Flask secret key (persisted across restarts)

Example:

BT_RESUME_PORT=8080 BT_RESUME_DEBUG=true btr serve

---

Requirements

• OS: Windows, macOS, or Linux
• Python: 3.10+
• RAM: 8GB+ recommended
• Disk: 2GB+ free space (for the AI model)
• Ollama: Installed automatically on first run

---

Tech Stack

Layer	Tool
Frontend	React 19 + TypeScript + Vite
Backend	Python 3.10+ / Flask
LLM Runtime	Ollama
AI Model	Mistral 7B (fast, efficient)
PDF Reading	pdfplumber
PDF Generation	ReportLab
Distribution	PyPI — pip install btr-resume

---

Building from Source

# Clone the repo
git clone https://github.com/rasumeng/BT-Resume.git
cd BT-Resume

# Install Python package in editable mode
pip install -e .

# Build the web frontend
cd web
npm install
npm run build
cd ..

# Start the server
btr serve

---

Project Structure

BT-Resume/
├── btr/                          # CLI package (btr serve, btr setup)
│   ├── __init__.py
│   ├── __main__.py               # python -m btr
│   └── cli.py                    # CLI commands
├── backend/                      # Flask REST API
│   ├── app.py                    # Flask app with CORS
│   ├── routes/                   # API endpoint definitions
│   ├── services/                 # Business logic (LLM, PDF, tailoring)
│   └── config.py                 # Backend configuration
├── core/                         # Shared Python modules
│   ├── prompts/                  # LLM prompt templates
│   ├── pdf/                      # PDF generation components
│   ├── resume_model.py           # Data models
│   └── utils.py                  # Utility functions
├── web/                          # React SPA frontend
│   ├── src/                      # Components, screens, styles
│   ├── package.json
│   └── vite.config.ts
├── pyproject.toml
└── README.md

---

Architecture

┌─────────────────────┐
│  React Web UI       │  ← Served by Flask / opened in browser
│  (User Interface)    │
└──────────┬──────────┘
           │ HTTP/JSON (localhost)
           ▼
┌─────────────────────┐
│  Flask REST API     │  ← Python Backend
│  (localhost:5000)    │
└──────────┬──────────┘
           │
           ▼
┌─────────────────────┐
│  Ollama LLM         │  ← Local AI Model (Mistral 7B)
│  (localhost:11434)   │     Runs entirely offline
└─────────────────────┘

---

License

MIT License — see LICENSE

---

Author

Built by Robert Asumeng LinkedIn | GitHub