Data scientist with three years of experience in applying machine learning, deep learning, statistical analysis, and data visualization techniques to solve real-world problems. Successfully completed several projects in domains such as financial technology and healthcare, using Python, SQL, and various tools and frameworks such as TensorFlow, PyTorch and Scikit-learn. I have a strong background in mathematics and computer science, with a bachelor's degree in computer science from Ain Shams University. I am passionate about finding insights from data and communicating them effectively to stakeholders and clients. I am always eager to learn new skills and technologies to enhance my data science capabilities.
- Developed a pricing optimization model leveraging time-to-sell metrics and historical data to balance market competitiveness, resulting in a 25% increase in adoption rates.
- Implemented a reinforcement learning-based consignment model using genetic algorithms to optimize buy prices for second-hand products while maintaining profit margins.
- Enhanced competitor price monitoring pipeline with automated scraping and data transformation capabilities.
- Created an intelligent product mapping system combining fuzzy matching, embedding similarity, and LLM-as-a-judge verification to accurately match competitor products to internal inventory.
- Developed a dynamic clustering take-rate model to segment products into optimized commission tiers based on sales velocity and margin profiles, replacing the previous fixed-commission per price-bucket approach.
- Optimizing marketing campaigns by segmenting customer base through advanced clustering techniques.
- Enhancing chatbot performance by integrating Retrieval-Augmented Generation (RAG) techniques, leading to more accurate and relevant responses.
- Developing a liveness detection model for fraud prevention, capable of distinguishing between real and spoofed images with high accuracy.
- Architecting a comprehensive taxonomy for ValU's merchant database, enabling intelligent search functionality within the app.
- Building a Recommender System to effectively match customers with relevant merchants, increasing engagement and satisfaction.
- Contributed to the development of an internal portal equipped with dashboards and analytics derived from the extracted metadata, empowering technical advisors to make data-driven decisions.
- Utilized BERT and topic modeling to analyze millions of research documents and extract valuable insights and research trends in the technology domain.
- Applied forecasting models to guide R&D, investments, and boost profits in areas like technology life cycle prediction, classification, and time series correlation and causation.
- Developed a model to extract entities and relationships between them, populating a knowledge base for taxonomy creation.
- Created a model to summarize transcripts, improving efficiency using a quantized Llama 2 model.
- Built an automated AI and GenAI Newsletter pipeline that generates statistical analysis of the collected papers and summarizes the most important papers and deploying it to be sent on a monthly basis to a set of direct recipients from the CTO to enable continuous monitoring of the AI research field.
- Spearheaded exploration of graph embeddings and pioneered application of Node2Vec technique within the team, enabling the discovery of hidden relationships and correlations among companies.
- Leveraged AutoGen for easy creation of next-gen LLM applications based on multi-agent conversations and integrating LangChain with it for time series narrative generation and summarization.
- Helped design Technical Interview Process & conducting interviews.
- Participated in preparing the annual hackathon conducted by Dell Technologies, targeting students and professionals covering areas as steganography, security, and machine learning.
- Conducted several workshops to demystify Data Science to Dell Technologies employees and hosting a competition for the trainees.
- Volunteered as a data science mentor for graduation projects for 5 different universities in Egypt under the AI Empower Egypt Initiative between Dell Technologies, MCIT (Ministry of Communications and Information Technology), and universities.
- Carried out in-depth study on O-RAN to grasp its structure, functioning, and future growth potential.
- Utilized Reinforcement Learning techniques to address complex problems in the network, such as load balancing, thereby optimizing network performance and efficiency.
- Stayed updated with the latest advancements in O-RAN and Reinforcement Learning to ensure the relevance and applicability of research work.
- Engaged in NLP tasks like Intent Classification, Slot Filling, and used OpenAi's GPT-3 API to construct a retail chatbot assistant.
- Developed an API using ONNX for real-time video matting, including support for processing input videos.
- Constructed an API for real-time video super-resolution operations.
- Developed an API by comparing Meta's BlenderBot2, SeeKeR, and BlenderBot3 models to use the most effective one for creating a precise chatbot.
