Stavan Shah

• Developed an AI-driven team matching agent to connect students with advisors, enhancing student engagement. • Led the website development utilizing React, Node.js, and SQL, streamlining user experience. • Collaborated with cross-functional teams to ensure seamless integration and functionality of the platform.

Graduate Teaching Assistant for the course: Senior Software Design. As a teaching assistant, I guide students through a CEN-focused project where we integrate knowledge from previous courses to tackle real-world problems. Throughout the semester, I help develop research and development skills while applying what students have learned in their undergraduate studies. We design and implement a significant software system or module, mirroring experiences in industry and government.

I contributed to the development of an analytics web dashboard centered around a geographical risk analysis package. This dashboard enhances flexibility by allowing users to customize parameter values through functions and displays the outcomes of habitat connectivity risk indices via graphical plots. Geohabnet aims to empower users in visualizing habitat connectivity risk indices based on their specific parameter inputs.

Led the development of a high-performance data processing pipeline with a focus on data engineering, AI-driven validation, and database optimization. Technologies: Python, pandas, psycopg2, PostgreSQL, Azure OpenAI, SQL, JSON, CSV Key Contributions: Data Engineering & Processing: Built advanced pandas scripts to handle large datasets, automating data cleaning, alias mapping, and deduplication for seamless integration. Database Architecture: Architected secure, scalable PostgreSQL database connections and custom ETL functions, optimizing data integrity and query performance. AI Integration & Classification: Leveraged Azure OpenAI for dynamic data classification and validation, significantly enhancing data accuracy for entities like cities, ports, and countries. Automation & Optimization: Implemented high-efficiency algorithms, logging, and data status tracking, reducing processing time and ensuring robust pipeline reliability. Collaboration & Documentation: Delivered maintainable, well-documented code and led integrations across a multi-functional team to achieve streamlined data operations.

As an SD2, I led development for a clinical trials search engine at Amgen, building scalable JavaScript RESTful APIs and deploying them on AWS (EC2, ECR, S3). I implemented a custom full-text search with advanced filtering and ranking, reducing query response time to under 300ms. I architected modular, object-oriented components that accelerated feature rollouts by 30% and streamlined onboarding. I also designed RESTful microservices, containerized with Docker and Kubernetes, and optimized CI/CD workflows using Jenkins. In parallel, I mentored junior developers on secure coding, performance tuning, and clean code practices—driving cross-functional delivery of end-to-end solutions.

As an SD1, I contributed to high-impact data-driven platforms for clients like BMS and Pfizer. At BMS, I led the end-to-end development of a Django-based forecasting tool using ARIMA, LSTM, and XGBoost—driving $2M in revenue. I deployed 20+ cloud-native microservices using AWS Lambda, API Gateway, and DynamoDB, and conducted performance profiling and end-user testing to ensure system reliability. At Pfizer, I redesigned a Python-based segmentation model using clustering algorithms, reducing operational costs by $1M. I scaled the platform for 10K+ users with multithreading, async processing, and Redis caching, while automating data workflows with SQL and Snowflake to improve reporting efficiency.

Pune Institute of Computer Technology

Devloped software connecting 500+ professionals offering grooming services to register users