Shahryar Khan

Nestlé Baby & Me (Jul 2024 – Present)

• Developed a recommendation system utilizing clustering techniques, deep learning, and artificial neural networks to identify popular customer products and provide targeted suggestions.
• Built an end-to-end analytical tool leveraging PandasAI, GPT-3.5/4, and system prompt engineering for business insights. The tool incorporated fine-tuned LLMs and Retrieval-Augmented Generation (RAG) to store questions and reusable code.
• Engineered the backend in Flask and the frontend in JavaScript, delivering a robust and dynamic user experience.
• Integrated a scalable pipeline for continuous updates and data retrieval, enabling real-time analysis and decision-making.

GSK MLMP Phase 2 (Sep 2024 – Dec 2024)

• Worked on clustering and deep learning techniques to optimize the discovery and development of medicine molecules.
• Designed and implemented LLM-based pipelines to extract critical data from documents, ensuring comprehensive information availability for biopharmaceutical models.
• Consolidated complex datasets into streamlined tables, creating a dynamic system for efficient data extraction and analysis.
• Leveraged existing GSK systems and introduced tailored attributes to enhance data accessibility and utility.

Vodacom South Africa (Oct 2022 – Aug 2024)

• Improved an existing recommendation system for CVM marketing campaigns using Naïve Bayes, Random Forest, and deep learning techniques to optimize package recommendations for customers.
• Enhanced SQL scripts for streamlined data retrieval and dashboard visualization, ensuring the most demanded KPIs were sorted and displayed efficiently.
• Created predictive scoring mechanisms and integrated them into marketing workflows to boost campaign effectiveness.
• Designed views and dashboards to simplify the analysis of response data for decision-makers.