-
-
landing page
-
input page
-
review
Inspiration
Job seekers struggle with ATS-optimized resumes. We wanted to democratize access to professional resume services while maintaining complete privacy through local AI processing.
What it does
Resume Builder Agent transforms existing CVs into job-specific, ATS-optimized resumes using local LLM models. It analyzes GitHub profiles, parses job descriptions, identifies skill gaps, and provides career coaching keeping the data privately on your machine.
How we built it
- Backend: Python with FastAPI, LangChain for document processing, Chroma for vector search
- Frontend: React with Vite for modern web interface
- AI: Local Ollama models (Llama 3.2) with RAG implementation
- Privacy: Complete offline processing with no external API dependencies
- Output: Professional PDFs using rendercv with YAML structured data
Challenges we ran into
- Balancing model performance with local processing constraints
- Creating a seamless dual interface (CLI + Web) experience
- Optimizing vector embeddings for resume-specific content
- Ensuring ATS compatibility across different resume formats
Accomplishments that we're proud of
- 100% local processing ensuring complete data privacy
- Intelligent skill gap analysis with learning recommendations
- Multi-format export (YAML, PDF, DOCX, LaTeX)
- Interactive AI agent that asks clarifying questions
- Full-stack implementation with modern tech stack
What we learned
- RAG implementation for domain-specific applications
- Local LLM optimization techniques
- Privacy-first AI system design
- Career coaching through conversational AI
- Balancing technical complexity with user experience
What's next for Resume Builder Agent
- Voice-driven resume building with speech-to-text
- Mobile app for on-the-go career management
- Industry-specific fine-tuned models
- Visual portfolio generation for creative fields
- Integration with job boards for automatic application optimization
Log in or sign up for Devpost to join the conversation.