Who I Am
A machine learning engineer specializing in advanced AI model development with a focus on cultural linguistics. I design and implement sophisticated neural networks that capture the nuances of regional dialects and deliver practical solutions to complex challenges. Driven by a passion for transforming cutting‑edge technology into impactful applications that bridge the gap between theoretical research and real‑world needs.
Skills
Programming
Web Development
Data Science & ML
DevOps & Cloud
Big Data
Web Services & DBMS
Experience
Machine Learning Engineer
August 2024 - PresentSawalni (Remote)
- Development of Darija Language Models: Led the design, development, and fine‑tuning of large language models specifically tailored for the Darija dialect.
- Data Collection and Analysis: Designed and implemented efficient data pipelines using Python and Pandas to collect and preprocess data for machine learning tasks.
- Pretraining and Fine‑Tuning LLMs: Pretrained and fine‑tuned large language models using PyTorch, Transformers (Hugging Face), and Langchain.
- Model Evaluation and Optimization: Employed tools like scikit‑learn and Hugging Face datasets for benchmarking and performance tuning.
Software Engineer
February 2024 - August 2024XcomSolution, Mohammedia
- Design and Develop of a Digital Marketing Automation.
- Microservices Architecture: Developed and implemented a robust and scalable microservices architecture.
- Full‑Stack Development: Contributed to both backend and frontend development, ensuring seamless integration.
- Cross‑Functional Collaboration: Partnered with business teams to align technical solutions with operational needs.
Junior Data Scientist
July 2023 - October 2023JPTRACK: JP&Co, Casablanca
- Developed Advanced Fuel Consumption Tracking Systems.
- Algorithm Design: Engineered and deployed custom fuel consumption algorithms optimized for various vehicle types.
- Theft Detection: Created machine learning models to detect and prevent fuel theft.
- Data Analysis: Performed detailed analysis to optimize fuel efficiency and reduce operational costs.
Volunteering & Open Source
Software Developer and Data Scientist
May 2024 - PresentAtlasAI
- Building the next generation of Moroccan AI Models.
- LLM Development: Creation and refinement of large language models tailored specifically for Darija.
- Data Collection Platforms: Designed and developed platforms to efficiently collect and preprocess data.
- Collaborative Innovation: Worked closely with a multidisciplinary team of data scientists and engineers.
Projects
Al-Atlas: Moroccan Darija Language Model
Developed Al‑Atlas, a 0.5B parameter language model, the first dedicated foundation model for Moroccan Darija, fine‑tuned from Qwen‑2.5.
TODa: Tamazight Open Dataset
Conceptualized and developed a groundbreaking open‑source project to preserve and advance the Tamazight language through an extensive linguistic dataset.
Tarjman-AI: Moroccan Chat-bot
Developed a multilingual question‑answering platform that allows Moroccan users to interact with advanced large language models (LLMs) in native languages.
EmbedPrepro: Text Analysis Library, CLI
Created a command‑line tool and library designed for text analysis tasks, including embedding, clustering, dimensionality reduction, and visualization.
Education
HASSAN II University, ENSET
2022 - 2024
Mohammedia, Morocco
Master's of Artificial Intelligence & Distributed Systems
IBN Zohr University
2021 - 2022
Agadir, Morocco
Bachelor's of Mathematics and Computer Science
Languages
Amazigh: Native
Arabic: Native
English: Advanced
French: Intermediate
