Skill set
Additional skills (37)
Zakaria in a few words
A Machine Learning Engineer, with over 2 years experience in the vibrant ecosystem of Abu Dhabi. A graduate of the inaugural batch of the prestigious MBZUAI with a Master’s in Computer Vision. Throughout these experiences, I have been able to contribute to the state-of-the-art in my field of study in Computer Vision, and professionally, it has enabled me to always be on the cutting edge of the newest innovations in LLMs, whether it’s fine-tuning, observability, RAG-powered or Agentic-based Applications. In order to productionize these innovations, my professional experience allowed me to grow familiar with a wider stack of tools such as Docker, Kubernetes, AWS Services and its configurations both through CLI and programmatically using tools like Terraform (e.g: S3, Redis, EC2, ECR, EMR, IAM, RDS, etc.).
◦ In my next role, I would like to be in an environment where I will continue to professionally grow more skills that
continue to set the trend and solidifying my skills acquired, whilst having a space to progress within the company into positions of more responsibility and management.
Experience
- Integrated Data IntelligenceAI & Machine Learning EngineerJanuary 2023 - Today (2 years and 5 months)Abu Dhabi - United Arab Emirates• Engineered and processed data from multiple sources, including web scraping and APIs, applying advanced parsing techniques, ML-based similarity calculations, and NLP models (RNNs, LLMs) for classification, sorting, and feature extraction.• Fine-tuned OCR models for a Conversational RAG agent, enabling secure document ingestion into a knowledge base without persistent access, leveraging retrieval-augmented generation and LLM fine-tuning. Developed and optimized the backend of the conversational AI system, implementing API design, user/memory management (confidentiality guardrails, geo-locking, access controls), and CI/CD pipelines for robust deployment and monitoring.• Played a key role in designing and productionizing the system architecture integrating OCR, LLMs, and RAG models, ensuring scalability, efficiency, and security. These tasks -including LLM fine-tuning and deployment- were accompanied by the best practices in terms of observability and model monitoring ensuring optimum production performance as well as accuracy and robustness.• Led critical components of a news analytics pipeline, handling data scraping, sorting, and deep learning-based relevance filtering, incorporating named entity recognition, summarization, and insight extraction models.• Skills and Tools: Python, Langchain, PyTorch, TensorFlow, Lightningand Scikit-learn for ML development, along with Docker, Kubernetes, and AWS services (EC2, S3, RDS, IAM) for scalable deployment. Employed FastAPI and Flask, and Asyncio for API design, Airflow and MLflow for pipeline automation, and Prometheus and Grafana for system monitoring and performance optimization.
- Capgemini EngineeringADAS R&D InternOctober 2018 - December 2018 (2 months)Casablanca, Morocco• Gained deep expertise in Neural Networks and their applications in Deep Learning for autonomous vehicle systems, focusing on real-time object detection.• Contributed to defining project scope and technical requirements, ensuring feasibility in both simulation and real-world testing environments.• Conducted an ablation study on state-of-the-art real-time object detection models, evaluating edge compatibility to determine the optimal architecture for the project's hardware and software constraints.• Improved the functionality and accuracy of an onboard camera-based real-time Object Detection model, optimizing performance for deployment.• Presented research findings and technical insights to department heads, team members, and client representatives, aiding strategic decision-making.• Skills and Tools: C/C++, Python, Matlab Scripting, VHDL (hardware description for FPGA design), Bash, nano, Simulink (model-based design and simulation), Numpy (numerical computing), Tensorflow (deep learning model development), Pytorch (deep learning and neural networks), Sci-kit Learn (machine learning algorithms), Wandb (experiment tracking and visualization).
- PricewterhouseCoopers (PwC)Technology Consultancy InternSeptember 2016 - December 2016 (3 months)Port Louis, Mauritius• Designed and developed a multi-functional dashboard using, which ranked among the top 5 most-used features in PwC's solution package for its largest client.• Contributed to regular penetration testing of delivered products, identifying and mitigating security vulnerabilities to enhance system robustness.• Developed a preliminary database management pipeline in MySQL, streamlining data access for the front-end team and improving workflow efficiency., and client representatives, aiding strategic decision-making.
Recommendations
Education
- Master of Science in Computer ScienceMohammed Bin Zayed University of Artificial Intelligence2023Masters of Science in Computer Vision;
- Bachelor of EngineeringGlasgow Caledonian University2020Bachelor of Engineering with Honours,