Karima Benzahra

Profile Master in Data Science and Computer Science Engineer. Worked on projects in OCR, medical video analysis, and SaaS platforms. Experience covers data analysis, backend development with Python and FastAPI, frontend development with Next.js and React, and DevOps with Docker, Kubernetes, GitHub Actions, and Jenkins. Focus on turning research ideas into usable applications that meet data sovereignty requirements. Experience Full-Stack Web dev | rankiha.com 2 Platform – Freelance • Built a web earning platform for Algerian users with FastAPI backend, Next.js frontend, and MongoDB. • Designed Points system with wallet, task rewards, withdrawals, and anti-fraud controls (IP/device tracking, screenshot verification). • Integrated Cloudflare for security and deployed production environment on Vercel (frontend) and Render (backend). • Implemented admin console for user management, task approvals, and withdrawal processing workflows. Remote Oct 2025 – Present Full-Stack & DevOps Engineer | Kalima OCR (Arabic Text Digitization SaaS) • Built the Kalima OCR web platform kalima-ocr.com 2 with Next.js (frontend) and FastAPI (backend). • Deployed services on a VPS using GitHub Actions with self-hosted runners, Docker, and Docker Compose for orchestration. • Implemented asynchronous processing with Redis and Celery for long-running OCR tasks. • Integrated two-phase LLM correction (via OpenRouter DeepSeek API) and contextual user chat. • Set up MinIO for file storage, MongoDB for metadata, and an authentication system with email verification (SMTP) and Google OAuth. • Developed custom monitoring and logging features to track requests, usage, and performance. Algiers Feb 2025 – Sep 2025 Co-Founder & Full-Stack Developer | RetroMuse Studios • Co-founded RetroMuse Studios, certified by EstinUP incubator with the national startup diploma (grade 19/20). • Developed and maintain the studio’s official website retromuse-studios.com 2. • Represented the studio at COSI’2025 (ESTIN), presenting its vision for Algerian game development. Web Development Intern | Biometrics Department • Collaborated on a web app for managing and retrieving official papers for the department’s secretary. • Built and integrated templates using HTML, CSS, JavaScript, Flask, and LaTeX. • Improved document management efficiency for official papers. Remote Dec 2024 – Present Tiaret Jul 2024 – Aug 2024 Education M.Eng. Artificial Intelligence & Data Science • Master Thesis: A Comparative Analysis of ML Deployment Approaches: A Case Study of Arabic OCR System • Engineering Thesis: Deploying Machine Learning Models: A Case Study on Historical Arabic Handwritten Text Recognition Preparatory Classes ESTIN Béjaïa Sept 2022 – Sep 2025 ESI Sidi Bel Abbès Dec 2020 – Jul 2022 Key Projects Kalima OCR Platform Digitized Arabic manuscripts with 95.7% success raet and ∼1s latency per line latency at kalima-ocr.com 2. Improved text reliability through twostage AI correction and enabled researchers to interact with manuscripts via contextual chat. 2025 Kalima OCR Monitoring Dashboard Developed a web dashboard to track OCR latency, AI correction, and user activity. Provided exportable reports and word-frequency analytics to support model refinement and research insights. 2025 Obesity Classification Model Predicted obesity levels with 95.6% accuracy using health data from 2,111 individuals. Applied preprocessing and compared models (Logistic Regression, Random Forest, KNN, SVM) with GridSearchCV tuning and cross-validation. 2024 Power BI Salary Analysis Built interactive dashboards analyzing global data science salaries across 2,000+ records. Delivered insights on salary trends by role, experience, company size, and remote ratio. Source: GitHub Repository 2. 2024 Skills Hard Skills: Python, JavaScript, SQL, FastAPI, Next.js, React, Tailwind, Flask, PyTorch, Scikit-learn, Docker, Kubernetes, GitHub Actions, Jenkins, Redis, MongoDB, MinIO, Power BI Soft Skills: Problem-solving, Teamwork, Communication, Research-driven mindset, Adaptability, Time management, Critical thinking Languages Arabic (Native) | English (Fluent, C1) | French (Conversational)