# Satvik P.

> Senior Full Stack Software Engineer | React, Node, TypeScript, AWS | US Citizen

Location: New York, New York, United States
Profile: https://flows.cv/satvik

Hi, I'm Satvik, a full stack React, TypeScript, NodeJS dev.

## Work Experience
### Senior Software Engineer @ Pantheon Labs
Jan 2020 – Present | Manhattan, New York, United States
- Created dashboard to view blood donation details for biotech Fortune 500 client with React, Javascript (TypeScript), NextJS, NodeJS (Express) and GraphQL with Kubernetes and CI/CD pipelines
- Led 5 person team to create dashboard for a VC funded startup with React (NextJS), Node (NestJS), Javascript (TypeScript) hosted on Google Cloud using Cloud Run and Postgres
- Developed full stack features for consumer ecommerce Fortune 500 client using microservices in Java / Kotlin and Spring Boot, with React, Javascript (TypeScript), React Query and NextJS for the frontend, hosted on Microsoft Azure using Azure VMs
- Redesigned admin dashboard used by hundreds with React, Javascript (TypeScript), NodeJS, Apollo GraphQL for Fortune 500 client, hosted on AWS using Lambda, with Cloudwatch, DynamoDB, AppSync, and Docker
- Led conversion of VC funded startup client app to Rust with GraphQL, Axum, WebSockets, Diesel ORM from NodeJS
- Led a 3 person team to create a voice/video/chat-based telemedicine web app where patients could talk to therapists for a subscription fee, using React, Node, Javascript (TypeScript), Twilio API, hosted on AWS with DynamoDB, for startup client
- Built a low-latency team productivity tool and chat application with Rust using Actix Web, Prisma ORM for backend REST endpoints and frontend in Flutter, Dart with over 1,600 people signed up
- Spearheaded transition from hard-coded marketing pages to Contentful CMS with React and Javascript (TypeScript) with a Python (Django) backend
- Built social network app for startup client with React Native frontend and Python (Django) backend
- Contributed to a real estate technology client’s project for scanning and mapping rooms using a Ruby (Ruby on Rails) backend and React and Vite frontend

### Software Engineer @ Booz Allen Hamilton
Jan 2018 – Jan 2020 | Washington, District of Columbia, United States
- Cut loading time of internal government website by 83% by redesigning it in Typescript, Gatsby, React, and NodeJS.
- Created React / Redux dashboard for analysts to view and flag cybersecurity threats, using an automated threat model with machine learning. This model achieved 98% accuracy, built in Python, Tensorflow and Keras to combat internal and external cybersecurity threats for the Department of Defense.
- Increased user efficiency by 50% for cybersecurity tasks by creating a chatbot that automatically loads in machine learning models and predicts threats using Python, Tensorflow, Keras, Docker and Kubernetes, deployed on AWS.
- Automated government OS and application security (RedHat Enterprise Linux, Postgres, Nginx, etc.) using Ansible and Terraform to cut down manual technician time from 2 weeks to 1 hour.

### Software Engineer Intern @ Immuta
Jan 2017 – Jan 2017 | College Park, Maryland, United States
- Enabled a 50% increase in data science speed by developing a machine learning model for differential privacy using Python, Tensorflow, Keras, Docker, and hosted on Google Cloud.
- Created AngularJS frontend for differential privacy model, which loads in under 100ms through frontend optimization.
- Integrated CSVs with Postgres to allow seamless execution of SQL queries on CSV files on multiple separate AWS servers with Apache Spark and Hadoop (HDFS), decreasing time for execution from 8 hours to 30 minutes.
- Increased Apache Hive and Impala query speed by 50% through meticulous analysis of performance metrics and bottlenecks.
- Presented to the 30 member team on the usefulness of machine learning by demoing algorithms and use cases.

### Software Engineer Intern @ Zentail
Jan 2016 – Jan 2016 | College Park, Maryland, United States
- Increased customer profits by 30% by building a custom machine learning based product repricer in PHP that learned from time of day, purchase type, and other factors, then integrated this into React as a full stack application and optimized page load speed to less than 100ms.
- Cut down on-boarding of new customers from 10 days to 2 days by creating a custom Naive Bayes Classifier in PHP that automatically categorized Amazon products using machine learning with an accuracy of 90% over 10k products. Integrated product classifier into React as a full stack application.


## Education
### Bachelor of Science (BS) in Computer Science
University of Maryland
Jan 2014 – Jan 2018

### Long Reach High School
Jan 2010 – Jan 2014


## Contact & Social
- LinkedIn: https://linkedin.com/in/satvikpendem

---
Source: https://flows.cv/satvik
JSON Resume: https://flows.cv/satvik/resume.json
Last updated: 2026-03-22