Chadlia El Amri

Software Engineer Intern at Choose Paris

Built AI automation tools reducing manual errors by 70% and accelerating data generation by 3x, processing 10K+ records daily. Replaced Google Sheets with an offer management platform, centralizing negotiations and automating contract workflows, reducing validation errors by 95% and manual processing time by 70%. Built scraping pipelines with rotating proxies, extracting 50K+ data points daily at 80% success rate. Delivered reusable packages following DDD principles, adopted by 4 teams and reducing code duplication by 40%.

R&D Intern at Latech Sousse

Conducted R&D on Lavague for automated booking systems, testing various agent configurations and optimizing the framework through custom prompts and workflow modifications achieving 70%+ task completion. Benchmarked 5+ LLMs (GPT-4, Claude, Gemini), reducing inference costs by 40% through prompt optimization. Contributed to full-stack AI application using LangChain for automated content generation workflows.

Software Engineer Intern at Latech Sousse

Conducted R&D on Lavague for autonomous web browser automation; evaluated GPT-4 vs open-source LLMs for complex web interactions. Engineered intelligent scraping pipeline handling 100+ sources with anti-bot mitigation, proxy rotation, and dynamic JS parsing. Architected RESTful APIs with NestJS implementing JWT authentication, rate limiting, and robust validation middleware. Delivered technical cost-benefit analysis comparing cloud vs self-hosted infrastructure for large-scale automation deployment.

R&D Intern at Latech

Conducted R&D on Lavague for automated booking systems, testing various agent configurations and optimizing the framework through custom prompts and workflow modifications achieving 70%+ task completion. Benchmarked 5+ LLMs (GPT-4, Claude, Gemini), reducing inference costs by 40% through prompt optimization. Contributed to full-stack AI application using LangChain for automated content generation workflows.

Software Engineer Intern at Choose

Developed full-stack application with React and FastAPI, processing 10K+ records daily with server-side validation; built intelligent automation workflows using LLMs, reducing manual data processing errors by 70%; replaced Google Sheets with a custom web interface and REST API, eliminating 95% of validation errors and supporting 10x throughput; implemented data processing pipelines with rotating proxies, extracting 50K+ data points daily at 98% success rate; deployed microservices on GCP Cloud Run with CI/CD pipelines, enabling zero-downtime deployments.