Ganesh Mahadev Udata

Enhanced questions generated by LLMs through prompt engineering and agentic AI. Developed a data pipeline to scrape news from 5 top RSS feeds and store in Firestore, integrating 100+ daily news updates into the Kwizit quiz game. Prompted the VertexAI model to generate current events questions based on data stored in Firestore, resulting in a 100% reduction in API utilisation costs for news integration.

Skills gained: Python, AI, Large Language Models (LLM’s), Software Development, Machine Learning, Statistical Analysis, Problem-Solving, Data Visualisation, Project Management, Communication. • Applied predictive modeling and large language models (LLMs) to analyze ITSM incident data, enhancing data driven decision-making, improving efficiency, and creating Power BI reports to visualize trends. • Developed a retrieval-augmented generation (RAG) application as a proof of concept, demonstrating how large language models (LLMs) can predict future incidents and recommend effective mitigation strategies. • Designed and implemented an LLM-based chat application utilising advanced prompt engineering techniques (priming/seed prompting) to enhance ITSM data analytics and facilitate more efficient incident resolution processes.

• Executed comprehensive analysis of ADMET properties for drug molecules, enhancing prediction accuracy of key pharmacokinetic behaviors. • Conducted extensive literature reviews to identify cutting-edge techniques for predicting Extraction properties. • Engineered XGBoost and Random Forest models, achieving a 20% improvement in predicting the Extraction property for novel drug candidates, accelerating the drug discovery process.

Skills: Object-Oriented Programming (OOP) · Statistics