Muhammad Usama Irfan

• Architecting and developing an enterprise-level data consolidation and analytics solution for United Industry Limited, enabling effective reporting and informed decision-making. • Designing and creating Analytical dashboards in Power BI for Teach for All, empowering the organization with valuable insights for informed decision-making. • Led a team of 3 developers in the development of an Intelligent Conversational AI Assistant for an EdTech company, leveraging semantic search and OpenAI LLM services.

• Engineered a Conversational AI assistant that enables users to discover precise information through instant messages and customized report views. • Improved performance by 10% by constructing a Natural Language Query Engine that converts text to DAX query. • Enhanced the user experience by introducing and developing new features such as Top-N queries, Lookup queries, Disambiguation, and Natural Language-driven scheduled & triggered alerts.

• Created solutions for a call center, focusing on sentiment analysis, NER, intent classification, and multi-label & multi-class text classification, achieving exceptional accuracy rates of over 87% for each task. • Implemented a cutting-edge Recommendation System for brand recommendations, leveraging GNN & Hybrid Recommendation techniques, which yielded remarkable scores • Spearheaded a team of 2 ML Engineers, overseeing the development and experimentation of models for the recommendation system.