Inspiration
Hiring is broken. Resumes and job titles often fail to reflect true skill or potential. We wanted to create a system that matches people to opportunities based on what actually matters — their abilities, not just credentials. Magnet was born from this frustration, a tool that understands people beyond LinkedIn profiles.
What it does
Magnet is a smart matching tool that helps companies find hidden talent by analyzing real skills, not just job history. It lets job seekers showcase what they can do and helps hiring teams identify overlooked but highly relevant candidates. Magnet surfaces potential, not just pedigree.
How we built it
- We used Bolt to build the entire MVP fast and efficiently. Our stack included:
- Supabase for authentication and database management
- Gemini API to generate career insights and enhance CVs
- OpenAI API to analyze user inputs (resumes, bios, skills) and intelligently match them to job requirements with ranked explanations
Challenges we ran into
- Figuring out the best way to extract and evaluate candidate skills meaningfully
- Prompt tuning to ensure the AI didn’t just regurgitate keywords but actually understood potential
- Balancing simplicity in the UI with the depth of insight we wanted to show
Accomplishments that we're proud of
- Built a functional MVP of Magnet in record time using Bolt, from idea to execution, we shipped fast without compromising quality.
- Implemented AI-powered matching logic that delivers ranked candidate recommendations based on real skills, not just job titles or keywords.
- Created a seamless one-click apply flow for employers, enabling faster hiring decisions and smoother candidate tracking.
- Designed a clean, usable recruiter dashboard that already supports intelligent match management and feedback loops.
- Sparked strong interest from early testers, especially among individuals struggling to find qualified but unconventional talent.
What we learned
- AI-based matching is powerful, but context matters, we had to experiment to avoid false positives
- Hiring bias can creep into anything unless you're intentional about how matches are made
- Fast feedback loops (build → test → refine) helped us progress quickly
What's next for Magnet
- Portfolio and project-based evaluation: We’re adding support for candidates to showcase real-world work, portfolios, projects, and demos, giving employers a deeper, more practical view of their capabilities beyond resumes.
- AI-powered recruiter dashboard: The recruiter dashboard is already live. We’re now enhancing it with smarter AI features to streamline decision-making, highlight standout candidates, and automate manual evaluation steps.
- One-click apply from external platforms: We plan to integrate Magnet’s one-click apply feature with third-party job boards, so candidates can apply seamlessly, while employers receive enriched skill-matched data within Magnet.
- Early access for hiring teams: We're launching a private beta with companies that prioritize skill-first hiring. This will help shape the next phase of the platform through real recruiter feedback.
- Empowering overlooked talent: We’re focused on making Magnet accessible to junior talent, self-taught devs, and career switchers, the ones traditional systems often miss, but who are capable of delivering great value.
Built With
- bolt
- gemini
- openai
- supabase

Log in or sign up for Devpost to join the conversation.