asser.mazin.20@gmail.com

• 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.