About FitScore
Why FitScore
Hiring shouldn't be a guessing game.
FitScore gives every candidate and recruiter transparent, mathematical job matching — so the best fit wins, not the best-formatted CV.
For candidates
Upload your CV once. Score any job description against it instantly. See exactly where you're strong, where the gaps are, and what you'd need to fix before applying — so you only chase roles you can actually land.
For recruiters
Paste or import a job description, upload a pool of CVs, and let the engine rank every candidate with evidence-backed match scores. No more sifting through hundreds of applications manually.
How it works
Your CV and the job description are first parsed by an LLM into structured data, then scored by a Python rules engine, then adjudicated by a second LLM pass that produces cited, dimension-by-dimension reasoning. You see exactly why you scored what you scored — so you can decide, not just accept a number.
What it is not
FitScore scores are probabilistic AI outputs — informational, not a hiring decision. They tell you why, so you can judge for yourself. Always verify with the original job description and use your own judgement.
Built with
On the Roadmap
Swipe-to-Apply
See your fit score, swipe right to apply instantly. Consumer-app speed for enterprise hiring.
Coming soon
Contextual Interview Prep
Based on your CV and JD, prepare technical and scenario-based interview questions tailored to your exact gaps.
Coming soon
The team
FitScore was built end-to-end with AI assistance throughout — from architecture to code review to testing. Below is a transparent account of who (and what) did what.
Darshan Singh Meel
Product & Project Manager · Main Architect
Product vision, architecture decisions, domain expertise in data engineering hiring, and every final call on scope and direction.
Claude Opus
AI Reasoning · Lead Architect
High-level system design, scoring pipeline architecture, schema decisions, and the matching logic specification.
Claude Sonnet
AI Chat · Lead Engineer · Web & UI Design
Primary implementation across backend, frontend, infrastructure, and UI design — the majority of code and visual design in this codebase.
Claude Haiku
AI Speed · Engineer · Bias Testing
Supporting implementation tasks and bias/edge-case evaluation of scoring outputs.
Gemini 2.5 Pro (Google)
Main Code Reviewer · Product Feedback
Independent code review passes, product critique, and cross-checking scoring logic against real-world hiring practice.
Gemini Flash 2.5 (Google)
E2E Testing · DevOps
End-to-end test runs, deployment checks, and infrastructure verification.
About the maker
Darshan Singh Meel
Senior Data Engineer & Architect · 22 years experience
I have spent over two decades designing and building large-scale data platforms — data warehouses, lakes, lakehouses, real-time pipelines, and analytics systems — across regulated finance, retail, and tech. Most of my career has been hands-on: Python, Spark, SQL, cloud-native data stacks, and the architecture that ties them together.
FitScore started as a personal tool to help myself evaluate job opportunities more rigorously. After applying many times to roles I was convinced I was a perfect fit for — only to be rejected without explanation — I wanted to understand exactly why. The platform is built primarily around data engineering and data analyst roles, with solid coverage of broader IT and software engineering positions. With some web app experience and a lot of help from AI pair programmers (Claude and Gemini), I turned it into a full product. The scoring engine draws directly on what I know about how data roles are evaluated in the real world.
Coverage & limitations
FitScore has been primarily tested on data engineering, data analyst, software engineering, and IT roles. It should work for other technical and professional roles, but our skill vocabulary may not cover every specialisation yet — and you may see a different score than expected. We're continuously improving coverage as more role types are tested.
FitScore is an independent project. It is not affiliated with any recruiter, job board, hiring platform, or AI provider. Job summaries and match scores are generated by AI and may contain errors — always verify with the original posting.
Some CVs and job descriptions used during development and testing were sourced from CW Jobs, Totaljobs UK, LinkedIn, and FreelancerMap.de. These were used solely for evaluating scoring accuracy and are not stored or redistributed.
Trademarks & attribution
Claude and Claude Code are products of Anthropic. Gemini is a product of Google. LinkedIn® is a registered trademark of LinkedIn Corporation. The Claude and Gemini icons on this page are custom representations and not official logos.FitScore is independently developed and not endorsed by Anthropic, Google, or any job board.
Built by humans + AI (Claude and Anthropic)
© 2026 FitScore