Data Scientist
Passionate data scientist with over 4 years of experience in academia and industry. I solve hard problems and bring value to business by harnessing the power of data analysis and machine learning. My toolkit includes Python, probability and statistics using numpy, statsmodels and pingouin, machine and deep learning using scikit-learn and keras, visualization using matplotlib and plotly, and I’m currently making video tutorials on YouTube to share some the knowledge and tips I have found useful.
Skills
- Machine learning (scikit-learn/tensorflow2) - Expert
- Python (pandas/numpy) - Expert
- Statistics (statsmodels/pingouin) - Experienced
- SQL (postgres/oracle) - Experienced
- Numerical optimization - Skillful
- Computer vision (opencv) - Skillful
- Git workflow - Skillful
- Linux - Skillful
- Time series analysis - Beginner
- AWS - Beginner
- NLP/LLMs/Chatbots - Beginner
- R - Beginner
- Kafka - Beginner
Work Experience
Data Scientist @ TAWAL - (August 2021 - Present)
Worked for TAWAL via three different outsourcing companies
Solutions by STC - Riyadh, Saudi Arabia - (August 2022 - Present)
- Achieved a 27% greater reduction in tickets generated from the Tower Management System by leveraging machine learning to identify associations between historical alarm data, surpassing the system’s existing rules
- Developed a Proof of Concept (PoC) to enhance technician ticket assignments by automating optimal
allocation, considering factors such as workload, geographical proximity, and skillsets, thus transitioning
from manual to automated assignment processes in the company
- Collaborated closely with the infrastructure team to facilitate a comprehensive overhaul of the company’s AI infrastructure and architecture which enabled the implementation of a complete MLOps lifecycle
- Developed a PoC QA chatbot that answers questions about the company’s knowledge base using LLMs, LangChain, and a Vector DB
Norconsult Telematics - Remotely - (January 2022 - August 2022)
- Developed a predictive maintenance application with a 72% outage detection rate for telecom sites, demonstrating advanced capabilities in outage prediction and prevention
- Effectively supervised Python-based ETL cron jobs for data extraction, transformation, and loading into destination databases, facilitating seamless visualization through BI tools mainly Qlik Sense
MasterWorks - Remotely - (August 2021 - January 2022)
- Applied analysis, machine learning, and statistical tests to diverse use cases including asset recommendation, performance evaluation, and replacement alerting, presenting findings to enhance operational efficiency for business
- Contributed to 3 impactful research projects, resulting in the publication of 4 conference papers, including 2 as the first author, covering areas such as human activity recognition using IMUs, converting hums to spoken words for speech-impaired individuals, and developing an indoor workout virtual coach.
Machine Learning Engineer @ DevisionX - Egypt - (October 2017 - November 2018)
- Executed computer vision tasks, achieving a remarkable 98% accuracy by training convolutional neural networks to identify Arabic/Hindi numerals in Egyptian identification cards, and also contributed to textile defect classification utilizing a combination of traditional and AI-based computer vision methods.
Education
M.Sc., Egypt-Japan University of Science and Technology - Egypt -(September 2019 - September 2021)
- Graduated with a distinguished GPA of 3.85
- Received a prestigious full teaching assistant scholarship
- Completed a rigorous curriculum encompassing Advanced Intelligent Systems, Advanced Machine Learning, Learning Algorithms, and Neural Networks
B.Sc., Faculty of Engineering, Mansoura University - Egypt -(September 2011 - September 2016)
- Graduated with Very Good with Honors
- Major in Computer Engineering and Systems
Projects
Data Scientist In-demand Skills Dashboard Link
- Scrapped around 400 data scientist job postings from LinkedIn
- Extracted job skills from the job descriptions using a pre-trained Named Entity Recognition model from HuggingFace
- Visualized the findings using Tableau and built an application using Streamlit to filter job skills requirements based on country, seniority level and other factors
- Made a series of videos explaining how it’s done Link
Inertial Gait-based Person Authentication Using Siamese Networks Link
- Proposed a Siamese Network-based framework for inertial gait-based person authentication
- The proposed framework allows for learning a model on a set of subjects and being tested on a totally different set of subject (authenticate new subjects) without retraining the model from scratch.