Tanvi Sharma

About Tanvi Sharma

I’m a Data Scientist passionate about transforming messy, large-scale data into actionable insights that shape smarter business decisions and user-centric product growth. With experience spanning financial risk analytics, anomaly detection, and predictive modeling, I’ve built and deployed production-grade, Python-based ML pipelines that reduce fraud risk, optimize credit decisions, and enhance user engagement.My work focuses on end-to-end system development from data exploration and modeling to validation and deployment using Python as my primary implementation language. I leverage AI-assisted engineering workflows using tools like VS Code, GitHub Copilot, and Cursor to accelerate experimentation, improve code quality, and maintain strong human oversight across complex systems.My toolkit includes SQL, Python, XGBoost, Bayesian networks, Airflow, and MLOps tools like MLflow/Kubeflow, but I thrive most when applying these to solve real-world product and financial challenges.I’m excited to connect with teams where data science directly drives outcomes, whether that’s reducing fraud loss, scaling product analytics, or building reliable, explainable AI systems for real-world use.

Experience

1. Co-founder

   Stealth Startup

   Aug 2025 — Present · Boston, MA, US

   Designing subtle intelligence for daily life- Developing and iterating on Python-driven AI systems using AI-assisted IDE workflows (Cursor), enabling rapid prototyping, controlled experimentation, and system-level validation- Architecting and building a next-generation AI system with an emphasis on human oversight, reliability, and scalable deployment.

Education

• Boston University

  Master of Science - MS, Applied Data Analytics

• Manipal University Jaipur

  B.Tech , Electrical, Electronics and Communications Engineering

  2017 — 2021