About Me
I started my career at Bank of Baroda, India's third-largest public sector bank, spending four years in credit analytics across 131 million customers, building risk models, segmentation strategies, and data-driven campaigns. It taught me how to make decisions when the stakes are real and the data is messy. But the more I worked with data, the more I was drawn to what it could predict, not just describe. That curiosity about predictive analytics and machine learning became impossible to ignore.
I made the leap and moved to the United States to pursue my Master's in Analytics at Northeastern University (3.93 GPA), where I immersed myself in statistical modeling, predictive analytics, and machine learning. I served as a Teaching Assistant for the Predictive Analytics course, redesigning the curriculum and running TA office hours. My growing fascination with deep learning and NLP led me to pursue additional coursework through Coursera's Deep Learning Specialization alongside my degree. It was the foundation I needed to go from analyzing data to building intelligent systems.
Fetch Rewards was where I truly came into my own. I joined as a Product Data Scientist in a fast-paced startup environment, diving into the data of 9 million active users, building KPI dashboards, performing deep-dive analyses, running A/B tests and experiments, and delivering reports and insights to senior leadership. What I valued most was the exposure to the full picture, not just building models in isolation, but understanding how data connects to product decisions, user behavior, and business outcomes.
I grew into a Tech Lead and Senior Data Scientist, leading a cross-functional team of engineers, data scientists, and MLEs. I owned the end-to-end delivery of Fetch's patented digital receipt information extraction system, driving it from initial concept through architecture, development, app integration, and production deployment. I handled stakeholder alignment, roadmap planning, sprint execution, and vendor replacement strategy. While leading the project, I stayed hands-on with the R&D, fine-tuning transformer models using Hugging Face, building semantic search systems with SBERT and FAISS, and shipping ML models on Amazon SageMaker. The in-house system replaced a third-party vendor, saving millions in operational costs and resulting in a patent. That experience confirmed what I already knew. I thrive when I have a holistic view of the problem and can connect the dots from the underlying technology to the end-user experience.
I took a deliberate career break for pregnancy and early childcare. During this time, I kept myself continuously learning, staying current with the rapid advancements in GenAI, LLMs, and agentic AI architectures.
Today, I'm back with renewed focus and deeper expertise in LLMs, RAG, and Agentic AI. I'm currently working as an independent contractor through Mercor on LLM evaluations and frontier AI red teaming, while actively seeking my next full-time opportunity, ideally at the intersection of technical leadership and product ownership, where I can drive end-to-end delivery of data and AI products. I'm a Green Card holder and do not require visa sponsorship now or in the future.