Mosaic & Me
Mosaic & Me is a skills-first hiring platform where candidates build structured capability profiles—their "mosaic"—that capture skills, projects, and evidence across seven key areas. The platform uses semantic AI matching with OpenAI embeddings and pgvector cosine similarity to align candidates with roles based on demonstrated capability rather than keywords. Candidate preferences like salary and location sharpen matches but remain private unless the candidate chooses to reveal them.
- HR Technology
- Artificial Intelligence
- Software Only
- Sales Technology
Funding
Founders
Founder details are not available yet.
Product
Problem
Traditional hiring relies on CVs and keyword-matching ATS systems that filter on job titles, degrees, and credentials rather than actual capability. This systematically undervalues generalists, career changers, and non-traditional candidates, while producing poor predictions of on-the-job performance and narrowing the talent pool unnecessarily.
Solution
Mosaic & Me is a two-sided, skills-first hiring platform where candidates build a structured capability profile—their mosaic—capturing skills, projects, and evidence across seven key areas. The platform uses semantic AI matching with OpenAI embeddings and pgvector cosine similarity to align candidates with roles based on demonstrated capability rather than exact keywords. Candidates set preferences around salary, location, and deal-breakers, which sharpen matches but remain private unless the candidate chooses to reveal them. Employers receive ranked shortlists of genuinely aligned candidates and can use a Calibration Sandbox to adjust requirement weights in real time and see the impact on the matched pool instantly.
Target Audience
Primary users are generalists, career changers, and non-traditional candidates seeking roles that match their full skill set, as well as employers and recruiting teams looking to implement skills-first hiring practices and reduce time-to-hire.
Features
- Semantic AI matching engine using OpenAI embeddings and pgvector cosine similarity to understand capability context, not just keywords
- Structured capability profiles that map experience across seven key areas with evidence and context attached
- Privacy-first design where candidate preferences and salary expectations are never shared with employers unless explicitly revealed
- Passive matching model that surfaces suitable roles to candidates without requiring active applications
- Calibration Sandbox for employers to adjust requirement weights and deal-breakers in real time and observe changes to the matched candidate pool
- Candidate preferences used to sharpen match accuracy while maintaining full control over information disclosure
- Career Explorer for adjacent-role discovery