Flagship experience · Atihmam Conseils
Atihmam ATS
Software Engineering Intern · Full-Stack & Backend
Built and maintain an AI-powered applicant tracking system that turns CV intake into a structured, explainable recruitment workflow—from document parsing and candidate scoring to recruiter decisions.
Reliable CV processing
PDF uploads move through an asynchronous Celery pipeline: SHA-256 hashing prevents duplicate processing, pdfplumber extracts text, and structured candidate data is persisted to PostgreSQL.
Multi-dimensional AI scoring
A seven-dimension scoring engine evaluates role fit, skills, experience, projects, and credentials with Google Gemini, backed by fallback LLM providers and recruiter-feedback-driven score adaptation.
Recruitment workflow automation
A six-stage Kanban pipeline applies state-machine validation to candidate transitions, while Google Drive webhooks keep document changes synchronized automatically.
Production-ready split architecture
React 19, Vite, and Tailwind CSS on Vercel connect to a Python 3.11 FastAPI service on Render, with Supabase PostgreSQL, Valkey/Redis, and Celery workers for durable background processing.
