Resume

University of Arizona

Master of Science in Data Science
GPA: 3.83/4.0

Shiv Nadar University

Bachelor of Technology in Computer Science
Two times Dean's List Awardee
Specialization in Data Science and Big Data Analytics

Amazon · Eller Immersion Project

Tiered Minute Reduction (TMR) Forecasting

Tucson, Arizona, United States

• Building a Tiered Minute Reduction (TMR) forecasting model on Amazon-provided synthetic data that predicts weather-driven delivery capacity loss at the region level 5 days ahead, capturing day-over-day snow accumulation that the current planning process treats as independent 24-hour windows buffered network-wide.
• Developing and comparing two model architectures, a distributed-lag logistic regression and a gradient-boosting ensemble (XGBoost, LightGBM, CatBoost) with monotonic constraints and SHAP feature attribution, evaluated by walk-forward backtesting grouped by storm event and selected on calibration and accuracy lift at the 5-day horizon.
• Designing a simulation of the Return to Station (RTS) re-induction cycle to test hold policies that keep undelivered packages from inflating next-day demand during multi-day weather events.

Biosphere 2 · AI/ML Software Engineer

GreenThumb Cloud Control System, Freight Farms

Oracle, Arizona, United States

• Built a sensor data-quality validation layer with Oracle SQL over the GreenThumb Cloud Control System's 26-table schema, classifying 58 sensors' readings against agronomic set points and dead-band ranges to flag values the inherited pipeline missed, testing locally via Docker Compose (Oracle 21c XE, Spring Boot, Python simulator) against 10,000+ real readings before validating against the live Amazon RDS Oracle schema via DBeaver and JDBC.
• Secured edge access to the in-farm Raspberry Pi over Tailscale VPN and SSH, tracing the telemetry path (SSE ingestion, SQLite buffering, JSON POST to a Spring Boot backend on AWS ECS Fargate) and mapping the sensor-to-database schema.

HCL Technologies · Software Engineer

Generative AI Development Team

Noida, Uttar Pradesh, India

• Built a RAG assistant over 4,000+ SharePoint and ServiceNow files (Qwen3-Embedding, pgvector hybrid BM25 search, Qwen3-Reranker, heading aware 512 token chunks), lifting top 5 hit rate from 74% (dense only) to 89% on a 935 question labeled set and cutting daily service desk tickets 26% (2,300 to 1,700).
• Developed an MCP server for automated incident lookup, priority escalation, and ticket reassignment, reaching 94% tool selection accuracy across 120 golden evaluation tasks and cutting resolution time 76% (105 to 25 min), shipped via GitHub Actions CI/CD with zero downtime and unit and eval regression suites.
• Deployed employee cost models over SharePoint data (Python, pandas, FastAPI, Azure SQL) predicting cost within 7% of actual (MAPE), running as a weekly automated workflow on Azure App Service via Docker and Azure DevOps CI/CD.
• Built Sales KPP (React, C#, .NET, MS SQL Server) to track employee sales targets against revenue and annual deal closure metrics, replacing manual Excel consolidation for HR, PMO, and L2 leadership, then deployed it to production with technical documentation.

Carrier · Web Developer Intern

Gurugram, Haryana, India

• Designed and developed the Account Payable Management System and Ticket Tracker for the sales team using HTML, CSS, JavaScript, jQuery, and jQWidgets.
• Integrated an FAQ chatbot that cut support queries 38%, from 450 to 280 per week, and added automated tests.

PricewaterhouseCoopers · Technical Intern

Gurugram, Haryana, India

• Extracted and parsed project metadata from CNCF repositories with Python and REST APIs, then built a graph database model in Neo4j mapping dependency relationships across cloud native projects.
• Correlated and visualized the data in Golang backed dashboards running in Docker, surfacing insights for the Advisory team.

Reflex — Collision Risk DetectionSource Code

Python, PyTorch, YOLO11, OpenCV, FastAPI, React, TypeScript, PostgreSQL

• Developed a near miss detection system for traffic video with YOLO11, ByteTrack, and OpenCV, tracking road users at 94.4% vehicle detection F1 against the Urban Tracker benchmark.
• Scored collision risk at 87.6% precision on hand labeled events, optimized with TensorRT for 3.8x faster inference (12 to 46 FPS), and built a dashboard mapping close call clusters by intersection.

RadarMD — Chest X-ray TriageSource Code

Python, PyTorch, MONAI, timm, Grad-CAM, ONNX, FastAPI, Docker, GCP

• Developed a multi-label classifier for 14 thoracic pathologies on chest X-rays, training DenseNet-121 and ConvNeXt with PyTorch Lightning on NIH ChestX-ray14 (112,000+ images, 30,000 patients) to 0.84 mean AUROC, a 6 point gain over baseline.
• Validated Grad-CAM localization against 880 ground truth bounding boxes, held missed diagnoses under 8%, and deployed via ONNX export (4x faster CPU inference, ~120 ms/image) to a Gradio app on Google Cloud Run with CI/CD.

Self-Correction for Human Parsing ModelSource Code

Python, Git

• Benchmarked the SCHP human parsing model against state-of-the-art methods across Look into Person (LIP), Active Template Regression, and Pascal Part, addressing label noise challenges.
• Achieved 84.7% mIoU on the LIP dataset, 2.3% higher than the result reported in the original SCHP paper.

Presented at the 35th International Conference on Database and Expert Systems Applications (DEXA) 2024, Naples, Italy

Graduate Grader, Computational Modeling in Cognitive Science, University of Arizona (Spring 2026).

Specialization in Data Science and Big Data Analytics, Shiv Nadar Institute of Eminence (May 2024).

Dean's List Award as a top 10% performing student at Shiv Nadar Institute of Eminence (Fall 2020, Spring 2021).