Apoorva Banubakode Email and Phone Number

2023-present : Focusing on enabling Generative AI Experiences- Ideate and develop end-to-end custom GenAI Tools and Agents as a part of a central capability team.- Research SOTA prompt engineering techniques, knowledge to then feed into above mentioned GenAI Tools catered to the specific LLM being used.- Led Evaluation effort for the custom GenAI Tools to track performance, efficiency, behavior amongst other metrics. Ensure benchmarking of custom Tools across various 3party Large Language Models. June 2021 -2023 : Traditional ML | Confidence / Uncertainty Estimation- Build ML models for TurboTax (via Express Experience), enabling a cohort of simple filers to reduce filing time from ~70+ mins to under 10mins. (https://www.intuit.com/blog/innovative-thinking/intuit-ai-drives-taxpayers-through-turbotax-express-lane-experience/ )- Explored uncertainity quantification techniques to gauge of predictions being provided. - Querying, Visualizing, Analyzing large amount of data (in Millions) and transforming them into usable datapoints for training an ML model.

My responsibilities include research and development of ML algorithms for financial applications. Patented an approach that helped small business owners make more informed decisions about their business.

Over these 2 years, I worked on:- Analyzing large scale data to better understand customers,- Propose solutions to automate manual tasks,- Develop POCs projects and experiences in the domain of Marketing Attribution, Root Cause Analysis of customer complaints, Forecasting, using Machine Learning to provide personalized solutions and insights.

Developed a library for desktop and Android Mobile developers to securely store confidential data. Implemented specifications of OASIS' KMIP - Key Management Interoperability Protocol (in Java) and used IBM's Security Key Lifecycle Manager to manage cryptographic keys that were used for encryption.