Karim Damak

— Design and deployment of a RAG system (LangChain + Chromadb) for precision agriculture, reducing document search time by 60% for 25+ users — Optimization of spectral regression models (+20% accuracy) via feature engineering and Bayesian tuning — Development of a full-stack Flask application and 3 secure REST APIs (FastAPI, <200ms, integrated monitoring) — CI/CD automation (GitHub Actions, pytest tests, Docker multi-stage): deployment reduced from 3 days to 2 hours — Collaboration with field teams and management to transform business needs into concrete AI solutions — Technologies: Python, scikit-learn, LangChain, OpenAI/Mistral LLM, Hugging Face, Chromadb, Flask, FastAPI, Docker, GitHub Actions

— End-to-end NLP pipeline: transformation of emails into voice messages (RNN TTS, 500+ emails/day) — ML industrialization with MLflow (tracking, registry, serving): rollbacks <5min — Deployment of Django REST application with web interface, reducing processing time by 40% — Good MLOps practices: reproducible pipelines, artifact versioning, drift monitoring — Python, NLP, RNN, MLflow, Django REST, Docker, PostgreSQL

— Django + PostgreSQL dashboard for network monitoring (10K+ calls/hour) — Implementation of RBAC, email/SMS alerting, and real-time KPI reporting — Network supervision script automation (-70% investigation time)