Sumanth Pandiri

|

Capgemini Invent

Driven to create solutions that serve people. My work spans full stack tools, AI/ML applications, robotics, solution engineering, and tech and strategy consulting.

About Me

I'm Sumanth, a tech professional currently based in Georgia. Throughout my career, I've developed and shipped full-stack web applications from the ground up for Air Force contracts, contributed to AI/ML projects, robotics projects, and held leadership in student organizations and community service organizations. I'm always on the lookout for opportunities where I can make a strong impact, grow my business and leadership skills, and take ownership of my work.

Featured Projects

IDEA Lab ML Project

IDEA Lab ML Application

As the sole developer as part of the Clemson IDEA Lab (now Vanderbilt), I created a web application for elementary students to learn AI concepts, such as image recognition, data biases, hyperparameter tuning, and train/test phases. This application was built using MobileNet, with the machine learning done entirely on the client side using TensorFlow.js. Students can train and test in live time, use their live camera or upload files, download and reuse models, and customize training parameters.

JavaScriptSvelteTailwind CSSTensorFlow.jsImage RecognitionTransfer LearningConvolutional Neural Networks

Key Achievements

  • 99%+ Model Accuracy, Live Testing
  • Minimal Training Data Required
  • Real-time, Speedy Processing
  • Customizable Features
View on GitHub
AWS Architecture Diagram

Amazon Web Services Capstone Project

As part of a team of 6, I architected an integrated user system using AWS Cloud Computing Services in a consulting project for our client, Clemson Makerspace. This system unified scattered means of data collection into a centralized data hub, eliminated lengthy, preliminary forms for their users, reduced costs, and collected valuable user demographics for client stakeholders. We incorporated a 3D printing digital interface using Raspberry Pis and OctoPrint, allowing for virtual printing and control, live machine monitoring and management, and automated data collection and storage through Webhooks, APIs, and AWS API Gateway, Lambda, and DynamoDB.

Amazon Web ServicesData ArchitectureAutomationREST APIsCloud Computing ArchitectureRaspberry PiPythonConsultingSolutions Architecture

Key Achievements

  • Significant Cost Reductions
  • Automated Data Collection
  • Live Machine Monitoring and Control
  • Centralized Data Hub
  • Simplified User Experience
KDMA Analytics Platform

Human-Inspired Multi-Agent Navigation using Knowledge Distillation

Developed during Summer 2023, this project is a kinematic, differential drive implementation of KDMA for multi-agent navigation. It implements a neural network trained on human movement patterns, as well as a reinforcement learning policy that minimizes colisions and maximizes efficiency. This was applied to TurtleBot4 with HTC Vive 2 trackers for real-time location and orientation tracking, using ROS/ROS2. The system demonstrates the potential for human-inspired navigation in multi-agent settings.

PythonROS2ROSTurtleBot4HTC Vive 2Autonomous NavigationMachine LearningReinforcement LearningRobotics

Key Achievements

  • Autonomous Navigation on TurtleBot4
  • Real-Time Location Tracking with HTC Vive 2
  • Kinematic Differential Drive Implementation
  • ROS2 Integration
  • Simulated Real-World Environment
View on GitHub
CERM Foundation Website Screenshot

CERM Foundation Website

The CERM Foundation is a non-profit organization that builds, serves, and celebrates underserved communities in Pickens County, SC. I was tasked with designing a website for the organization that would be used to showcase their mission, programs, and events. I corresponded directly with the organization's leadership to ensure the website was a true reflection of their values and mission. I am still involved with the organization, and have been able to help them with their digital needs.

WixWeb DesignConsultingUser ExperienceOrganization Brand and Image

Key Achievements

  • Complete Website Design
  • Portrayed Foundation Branding, History, and Mission
  • Search Engine Optimization (SEO)
  • Event Management and Past Event Showcase
Visit Website

Skills & Technologies

Programming Languages

Python
C#
C
C++
SQL
JavaScript
TypeScript
HTML
CSS
Java

Frameworks & Libraries

ASP.NET (C#)
Svelte
React
Next.js
Node.js
Bootstrap CSS
Tailwind CSS
Wix

Infrastructure & DevOps

Amazon Web Services
Google Cloud Platform
Git
Linux
Oracle
SQL Server
REST API Development
Serverless Architecture
Vercel

Robotics

ROS2
ROS
Raspberry Pi
Arduino
Turtlebot
Computer Vision
Motion Planning

Artificial Intelligence & Machine Learning

TensorFlow
PyTorch
TensorFlow.js
Pandas
NumPy
Matplotlib
Scikit-learn
CNN
LSTM
RNN
Deep Learning
Reinforcement Learning
Model Interpretability
Transfer Learning
Computer Vision

Soft / Leadership Skills

Leadership & Initiative
Cross-Functional Collaboration
Communication
Project Ownership
Agile
Technical Writing
Consulting
Solutions Architecture
System Design

Let's Connect!

Atlanta, GA