Experience

1. Feb 2025 - Present, New York - NY

   Peloton Interactive, Inc.

   Machine Learning Engineer

   • Owned and productionized an agentic AI instructor using LangGraph, enabling multi-step, guardrailed LLM workflows for personalized training plan generation.
   • Shipped onboarding personalization using GPT-4 and XGBoost to capture user mood and intent, improving early-session engagement and reducing cold-start friction.
   • Improved ranking and retrieval quality via transformer-based candidate generation combined with Neo4j graph signals, driving measurable MAP@K gains.
   • Led a cross-service redesign of plan and insight invalidation pipelines (Kafka, DynamoDB, gRPC), improving data correctness, system reliability, and long-term maintainability.
   • Raised team velocity by standardizing evaluation, monitoring, and error-handling across personalization services, while also driving cloud cost savings through infra cleanup.
2. Jun 2024 - Dec 2024, New York - NY

   Barclays

   Data Scientist Co-op

   • Automated quality validation for 10K+ datasets, improving efficiency.
   • Developed a LLaMA-based metadata analysis tool, ensuring $100K savings.
   • Streamlined Monthly Business Review process, reducing analyst effort by a week.
3. Sep 2021 - Jul 2023, Karnataka - IN

   Tata Consultancy Services

   AI Engineer

   • Accelerated blood sample analysis to 800ms using YOLOv4, saving $400K annually.
   • Developed catheter inspection system with Meta Detectron2, increasing precision by 25%.
   • Built tools to convert 2D CT scans into 3D visualizations, enhancing diagnostics.
4. Nov 2020 - Sep 2021, Karnataka - IN

   Data Engineer

   • Deployed a 40TB AWS data lake with ETL pipelines for Genentech.
   • Developed a sales analysis tool generating $2M+ in revenue within one quarter.
   • Reduced storage costs by 90% through tiered data distribution.
   • Led research in NLP-based report summarization and employee recommender systems.
5. Jun 2019 - Oct 2019, Punjab - IN

   Thapar University

   Data Science Research Intern

   • Published Adaptive Ensemble Model for COVID-19 diagnosis with 98% accuracy.
   • Discovered genetic biomarkers for Multiple Sclerosis using feature analysis.
   • Introduced a TensoPIT Bloom Filter with Kalman forecasting for real-time data caching.
6. Sep 2023 - Present, New York - NY

   NE Big Data Hub

   Graduate Student Assistant

   • Led the Covid Information Commons team on pandemic policy response inference.
   • Developed dashboards and visualizations to advance data science research.