NextRole
NextRole is a GenAI-powered career advancement tool that automates interview preparation. It takes your CV and a job description, then uses a multi-agent system to produce a tailored resume PDF, an interview prep document, and a day-of battlecard cheat sheet. It is for developers and professionals who want to streamline job applications and interview prep.
🎬 NextRole: AI Agents That Tailor Your Resume & Prep You for Interviews (Demo) · Tam Nguyen
✨ Key features
- Tailored resume PDF via rendercv
- Deep company and role research
- Structured interview prep with STAR stories
- Day-of battlecard PDF cheat sheet
- Time-boxed prep plans (1 month, 2 weeks, 3 hours)
- Multi-turn chat for iterative edits
🎯 Use cases
- Tailor your resume for a specific job application
- Prepare for an interview with company research and coaching
- Create a quick reference cheat sheet for interview day
- Plan your interview prep over a specific timeline
📦 Installation
🧰 Requirements: Requires Docker, API keys for OpenAI, Tavily, and LlamaCloud (or alternative providers), and optionally AWS credentials for Bedrock.
git clone https://github.com/tam159/next-role.git
cd next-role
cp .env.example .env # then fill in your API keys (see table below)docker compose up -ddocker ps # read the 0.0.0.0:<host>->... mappings🚀 Usage
# After installation, open the frontend UI and follow the in-app instructions.
# Upload your CV and a job description (or JD URL), set your prep timeline, and let the agents work.
# You can also configure your LLM models in the app's Configuration dialog.
❓ FAQ
What do I need to run NextRole?
You need Docker, API keys for OpenAI, Tavily, and LlamaCloud (or alternative providers), and optionally AWS credentials for Bedrock.
Can I use free or local LLMs?
Yes, you can use Google AI Studio's free tier for Gemini models, or point OPENAI_API_BASE to LM Studio or Ollama for fully local models.
How do I change the LLM model?
Open the in-app Configuration dialog and set the main agent and subagent models to a provider:model string (e.g., anthropic:claude-sonnet-5).
What is the shell-command approval feature?
Risky execute commands pause in the UI for approval; a read-only allowlist auto-approves trivial commands. It is on by default and can be disabled via CAREER_AGENT_EXECUTE_APPROVAL=false.
📊 Repository
🤖 Overview, features, install steps and FAQ were generated from the project's README on Sep 4, 2026. Always check the original source before running commands.