Hiberus IT

Machine Learning Engineer
July 2024 - Present

  • Designed and implemented a scalable Retrieval-Augmented Generation (RAG) system, enhancing information retrieval by 30% and reducing search latency to sub-second levels for enterprise knowledge management in specialized domains (Industrial, Media, Government, etc.).
  • Fine-tuned state-of-the-art Large Language Models (LLMs) including LLaMA, Phi, Mistral, and BERT for NLP tasks, achieving 95% accuracy while reducing training costs by 40% through low-bit quantization and efficient parameter pruning.
  • Developed an Autonomous Multi-Agent AI framework that automated customer support and data analysis workflows, cutting manual intervention by 50% and improving response times by 60%.
  • Led the technical design and architecture of multiple AI projects, ensuring 100% on-time delivery while enhancing system scalability and performance.

Machine Learning Engineer - Internship
Jan 2024 - July 2024

  • Designed and deployed a hybrid RAG system, integrating sparse and dense embeddings to improve context retrieval, reducing query resolution time by 35% and boosting user satisfaction.
  • Developed AI proof-of-concepts (PoCs) for enterprise clients, surpassing baseline performance metrics by 25% and securing multiple new business contracts.

Technologies:
Hugging Face, LLM, LangChain, LLamaIndex, Transformers, PyTorch, Vector Stores, OpenAI, Azure Cloud, AWS, FastAPI, Docker, Streamlit

MARSA MAROC Casablanca

Data Science - Internship
Aug 2023 - July 2023

  • Developed a Vehicle License Plate Detection system with 97% accuracy, reducing manual verification workload by 60% and improving port security and logistics tracking.
  • Created a Django/Angular real-time monitoring dashboard, streamlining data retrieval and enabling non-technical staff to generate reports 2x faster, improving operational decision-making.

Technologies:
OCR, Fine-tuning, Deep Learning, Angular, Django, AI, Object Detection

Technocolabs Softwares Inc.

Data Science - Internship
Feb 2023 - Dec 2022

  • Developed ML models predicting mortgage prepayment risk with 82% accuracy, enabling financial institutions to refine risk assessment and reduce default rates by 20%.
  • Built interactive financial risk analysis dashboards, helping stakeholders optimize portfolio strategies and improve investment decision-making by 30%.

Technologies:
Scikit-learn, Pandas, NumPy, Machine Learning, Flask, Docker, AWS

Ministry of Energy Transition and Sustainable Development - Kingdom of Morocco

Computer Science - Internship
Sep 2022 - Aug 2022

  • Developed a digital internship management platform, automating workflows and reducing processing time by 60%.
  • Designed an optimized database and real-time dashboard, improving administrative efficiency in resource planning and reducing application processing delays by 42%.

Technologies:
PHP, HTML, CSS, JavaScript