I’m a Master’s student in Computer Engineering at NYU Tandon, building production-grade systems at the intersection of Machine Learning, Software Development, and Cloud.
Most recently at McKinsey & Company (MIO Partners), I architected a serverless data lake using AWS Glue (Apache Spark), Iceberg, and Athena that cut an Excel-extract workflow by 58% and reduced data fetch latency by 70%.
Previously at Cimpress India, I migrated to a high-performance pricing architecture by restructuring .NET and Node.js service-level code, optimizing APIs to sustain 150k+ RPM and improving end-to-end latency by 25%.
- Architected and implemented a serverless data lake solution using AWS Glue (Apache Spark), Iceberg, and Athena, resulting in a 58% reduction in the average time taken to run the TradeOps data retrieval cycle.
- Redesigned the Trades OT extract pipeline using Parquet columnar storage and Presto parallelized execution, decreasing data fetch latency by 70% while enabling dynamic schema adaptability for real-time field additions.
- Led the migration to a high-performance pricing architecture by restructuring .NET and Node.js service-level code, implementing new functional features while optimizing database I/O, reducing read/write latency by 30% and supporting 150k+ requests per minute.
- Engineered full-stack pricing workflows using React, GraphQL, and Express, achieving a 25% reduction in end-to-end latency and ensuring consistent page loads across 3 global regions.
- Developed and rolled out business-critical pricing tools utilizing AWS, Docker, and Kubernetes to streamline infrastructure and cut deployment times by 40%.
- Collaborated with product designers to refine component state logic and API contracts, enhancing frontend reliability and accessibility.
- Championed code reviews focusing on scalability/security in Python/Node.js, reducing production incidents by 20%.
- Directed 90+ students (30 sophomores, 60 freshmen) in the web development domain, overseeing end-to-end project lifecycles from initial planning to deployment.
- Spearheaded backend development for Trailer 2 You, a trailer-booking app, utilizing Node.js, Express, and MongoDB.
- Optimized performance by implementing efficient database structures, reducing data retrieval times by 20%.
- Configured AWS Elastic Load Balancers and Auto Scaling Groups to sustain requests exceeding 10k per minute.
- Engineered a complete booking process from React UI to Node.js backend, integrating Socket.io and Google Maps API.
Designed a comprehensive ML system automating extraction and indexing of enterprise multimedia data. Integrated a multimodal pipeline using Whisper for transcription, Sentence Transformers for semantic search, and Flan-T5-Large for RAG. Optimized deployment via Ray and quantization to achieve <200ms embedding latency.
View ProjectArchitected the backend for a trailer-booking application (Uber-style) using Node.js, Express, and MongoDB. Engineered the full booking workflow from React UI to backend, integrating Socket.io for live map updates and Google Maps API to lower booking abandonment by 8%.
View ProjectDeveloped an automated music selection application based on real-time facial expression analysis. Built the entire UI and participated in the development of algorithms to map moods to playlists, streamlining user engagement through interactive ML features.
View Project