findr-ai: AI-Powered Candidate Screening Platform

Alt text

findr-ai is built to modernize technical recruitment by eliminating manual resume reviews, unreliable self-reported experience, and spreadsheet-based candidate tracking. The platform uses artificial intelligence to evaluate resumes, match candidates to job roles, and verify GitHub contributions with a level of depth that traditional applicant tracking systems cannot provide.

What is findr-ai?

findr-ai is an AI-driven hiring platform that modernizes technical recruitment by combining resume analysis, smart job matching, and automated GitHub project verification. Unlike traditional ATS systems that rely on basic keyword matching, findr-ai uses Claude Sonnet 3.5 with multi-agent CrewAI workflows to deliver comprehensive candidate insights in real time.

Key Features

  • AI-Powered Resume Analysis: Multi-crew CrewAI agents parse PDFs, extract skills, and match candidates to job descriptions with precise scoring Alt text

  • GitHub Verification System: Validates project authenticity using Firecrawl web scraping and Gitingest repository analysis

Alt text

  • Smart Job Matching: Claude Sonnet evaluates candidate-job fit with detailed scoring (0–100%) and hiring recommendations

Alt text

  • Real-Time Processing: Asynchronous analysis with live progress tracking and in-depth credibility reports Alt text
  • Dual Interface: Separate dashboards for companies (post jobs, review candidates) and applicants (track applications, view AI scores) Alt text

Multi-Crew Architecture

Crew 1: Resume Analysis

  • PDF Parser Agent extracts resume text
  • Resume Analyzer evaluates skills and experience
  • Job Matcher assigns a 0–100% suitability score
  • GitHub Extractor identifies repository links

Crew 2: GitHub Verification (always triggered)

  • Firecrawl Agent scrapes GitHub profile activity
  • Repository Analyzer uses Gitingest for deep code review
  • Authenticity Validator detects fake or inflated projects
  • Report Generator produces hiring insights and recommendations

Tech Stack

  • Frontend: Next.js 15 + React with Tailwind CSS and shadcn/ui (deployed on Vercel)
  • Backend: FastAPI + CrewAI orchestrating eight specialized AI agents powered by Claude Sonnet 3.5
  • Database: Supabase (PostgreSQL) for job listings, applications, and company profiles
  • AI Integrations:

    • Anthropic Claude Sonnet 3.5 for reasoning and scoring
    • Firecrawl API for GitHub activity extraction
    • Gitingest for repository and code analysis

Why findr-ai Stands Out

Traditional hiring platforms often fail to detect fake projects and exaggerated resumes. findr-ai overcomes this by verifying every GitHub repository, analyzing real code quality, project complexity, and developer authenticityproviding hiring teams with actionable intelligence instead of surface-level keyword matches.


Built With

  • 4.5
  • claude
  • sonnet
Share this project:

Updates