Othmane OUKBIL

A junior AI Engineer on a collaborative project between OCP and McKinsey, contributing to the development of a Generative AI solution designed to support the corporate strategic intelligence team in competitive analysis. – Analyzed and understood a large-scale codebase developed by McKinsey’s GAIL team. – Optimized data scraping processes for improved efficiency and performance using parallel computing. – Integrated AI agents tailored to client-specific requirements. – Completed five courses on Generative AI offered by McKinsey. – Participated in five learning sessions with international Generative AI experts from McKinsey. – Benefited from one-on-one coaching sessions with McKinsey experts during the entire internship period.

Delivered high-quality prompts that surpassed State-of-the-Art (SOTA) Large Language Models, enabling the model to generate superior responses in programming and mathematical fields. Ensured responses were not only technically accurate but also human-like, enhancing usability and user satisfaction.

Under the supervision of Pr. Younes OMMANE from the UM6P-FMS Lab, the project I applied ARIMA, LSTM, and RNN models to forecast drug resistance in 13,753 MTB isolates from 26 African countries. High-performance computing (TOUBKAL) was used for large-scale data processing, though prediction accuracy was impacted by sparse and uneven data distribution.

Developed a NLP-based preselection model, ”Robot-Preselector,” to optimize the recruitment process by recommending the best candidates to recruiters and the most suitable job positions to job seekers based on their resumes (CVs). Methods : – Used Selenium/BeautifulSoup for data scraping. – Preprocessed data from resumes and job offers, including PDF-to-text conversion and data cleaning. – Applied word embedding and used BERT to enhance the semantic understanding of text data. – Implemented cosine similarity to calculate the match between CVs and job offers. – Integrated Elasticsearch for efficient data storage and retrieval. – Developed a web interface using Streamlit for easy interaction with the model.

- Advanced databases. - Advanced Programming. - Deep Learning & Advanced machine learning . - Parallel Computing & HPC. - Natural language processing ( NLP ) - GPU programming ( CUDA ) - Graph on machine learning. - Numerical Analysis. - Operational research & Optimization. - Stochastic Optimization. - finite element/volume/difference. - Numerical analysis and EDPs. - Reinforcement learning. - Graph theory. - Physics informed neural networks. - Bioinformatic & Computational biology. - Dynamic of games and learning. - High dimensional approximation.

- Algorithmic & data structure and programming ( Python, Java, C & C++ ) - Operating systems ( Linux, CentOS , Ubunto and Shell programming ) - Data bases and NoSQL programming. - Data analysis , Data mining & Artificial intelligence. - Web development ( UML, PHP, Html, CSS, Web scrapping...) - Algebra ,Calculus & numerical analysis. - Probability & statistics. - Operational research and optimization. - Stochastic process. - Econometrics & Time series analysis. - Project management.