Data & Analytics Engineer · AI · Finance

Saumya Jain

I build data-driven tools and analytics systems that help financial teams make faster, smarter decisions. Currently an Application Engineer at Vanguard, finishing my CS degree at Drexel.

Saumya Jain

A bit about me

I got into computer science because I liked building things. I stayed because I realized the best things to build are the ones that help someone make a better decision, faster.

Right now, I'm an Application Engineer at Vanguard, where I work on infrastructure that supports teams managing hundreds of billions in assets. I've designed serverless GraphQL APIs, and built an LLM-powered research assistant on Amazon Bedrock that lets investment analysts query portfolio data through conversation instead of spreadsheets.

Before Vanguard, I did a data analytics co-op at Exelon, where I trained a YOLOv8 computer vision model that hit 91% accuracy for asset classification. Before that, I was a Data Analyst at Vibranic Global, using Python, MongoDB, and Tableau to find patterns in customer behavior that actually changed how the team made decisions.

The thread through all of it: I like problems where engineering and business context collide. The messy ones. The ones where the hard part isn't the code, it's understanding what the code needs to do.

When I'm not building, I'm usually shooting cars and street photography or watching Formula 1 and pretending I understand tire strategy.

3.71
GPA — Drexel University
3
Industry co-ops completed
91%
ML model accuracy — Exelon
UPE
Computing Honor Society member

Where I've worked

The Vanguard Group, Inc. Mar 2025 — Present
Application Engineer · Malvern, PA
  • Designed and deployed a serverless API using AWS Lambda, API Gateway, and OKTA authentication, enabling secure GraphQL data queries
  • Architected a full-stack LLM research assistant on AWS Bedrock integrated with Aladdin APIs and S3, enabling multiple investment teams to query portfolio analytics, time-series fund data, and ETF basket data through natural language
  • Implemented OAuth-secured API integration and built a serverless conversational interface, replacing manual data workflows and giving analysts on-demand access to research notes and fund insights
Exelon Corporation Apr 2024 — Sep 2024
Advanced Analytics & BI Co-op · Newark, DE
  • Applied computer vision and ML models (YOLOv8, scikit-learn) to automate asset classification, replacing a manual inspection process
  • Created automated ETL + data pipelines integrating SQL and Pandas, enabling continuous model retraining that improved accuracy from initial baseline to 91%
  • Partnered with analytics leadership to integrate outputs into Power BI dashboards, driving faster maintenance scheduling decisions
Vibranic Global Inc Jun 2023 — Dec 2023
Data Analyst · Princeton, NJ
  • Performed data analysis using Python and MongoDB to identify market trends and customer behavior patterns, surfacing insights that reshaped the marketing and supply-chain strategy
  • Implemented workflow automation via Monday.com, optimizing supplier onboarding and logistics efficiency
  • Collaborated cross-functionally to translate insights into strategy using Tableau and internal reporting dashboards
Confidence Clothes Youth Sep 2020 — Sep 2022
Director of Web Development · Princeton, NJ
  • Led the design and development of a responsive website using HTML, CSS, and JavaScript, improving user engagement
  • Managed the Web Development Committee, delegating technical tasks and cross-functional collaboration

What I've built

Vanguard RCS/SMS System 2025–2026

Built and tested a secure RCS messaging framework for Vanguard as part of a 6-person Drexel senior design team. The system replaces legacy SMS with encrypted Rich Communication Services for client-advisor appointment management. Developed the serverless backend on AWS integrated with Twilio's RCS API, with intent classification to parse free-text client responses and route real-time notifications.

AWS Lambda API Gateway DynamoDB SNS CloudFormation Twilio RCS Python
ARK 2023

A modular voice- and text-based assistant built with Flask and Python. Integrates APIs for real-time news, weather, web search, and YouTube content. Uses AJAX for async requests to minimize latency, with a real-time Flask UI and extensible plugin architecture.

Python Flask AJAX REST APIs Web Scraping

My toolkit

Languages
Python Java JavaScript/TypeScript SQL C HTML CSS
Frameworks & Libraries
React.js Flask Pandas NumPy Scikit-learn TensorFlow YOLOv8 Matplotlib
ML & Analytics
Computer Vision NLP / Intent Classification EDA Data Preprocessing Feature Engineering
Cloud & Infrastructure
AWS Bedrock Lambda S3 API Gateway DynamoDB SNS CloudFormation Azure ML
Developer Tools
Git VS Code JIRA Confluence Tableau Power BI Figma Twilio

Through my lens

Cars, streets, and whatever catches my eye. Replace these placeholders with your actual photos.

Let's connect