About
Hi there! I'm Sowmith Kuppa, an M.Sc. in Computer Science graduate from Old Dominion University, where my focus lies in machine learning, MLOps, and reliable AI systems.
I've built and deployed deep learning models in PyTorch and TensorFlow, designed Retrieval-Augmented Generation (RAG) pipelines, and developed data-driven experiments that improved model performance and research outcomes. My recent work combines AI reliability, cloud automation, and data engineering, turning complex research ideas into scalable, production-ready systems.
Before diving into AI, I worked in IT consulting and systems administration, where I managed large-scale Windows and Linux environments, automated infrastructure with Ansible and Terraform, and led CI/CD initiatives using Jenkins and Kubernetes. That background gives me a strong understanding of real-world system reliability and the importance of clean, automated deployment pipelines.
I'm passionate about building responsible, reproducible AI solutions, from data preprocessing to monitoring and model refinement and I'm always looking for ways to bridge research, automation, and practical impact.
Skills
Machine Learning & AI
PyTorchTensorFlowData Science & Analytics
Python
Pandas
NumPyMLOps & Automation
KubernetesDockerTerraformAnsibleJenkinsArgoCDHelmGitLabCloud & Containerization
GCPAWSKubernetesDockerDatabases
PostgreSQLMySQLProgramming Languages
PythonGoJava
JavaScript
HTML
CSS
Bash
C++Observability & Monitoring
PrometheusGrafanaCollaboration & Workflow Tools
JIRAExperience
AI Research Intern
Old Dominion University
January 2025 - Present (1 year)
Norfolk, Virginia, United States
- • Designed and implemented machine learning models using PyTorch to support research on reliable AI systems, focusing on model robustness and performance evaluation.
- • Developed and integrated Retrieval-Augmented Generation (RAG) pipelines to enhance dataset quality and context relevance, improving model accuracy and research insights by 20%.
- • Created data mining and drift detection algorithms to analyze large datasets, generating actionable visual reports to monitor and explain model behavior over time.
- • Performed iterative model refinement and testing, leading to measurable improvements in performance and contributing to findings on reliable and scalable AI techniques.
Graduate Research Assistant
Old Dominion University
October 2023 - May 2025 (1 year 8 months)
Norfolk, Virginia, United States
Key Responsibilities:
- Kubernetes: Hands-on experience with Jenkins job configurations, pipelines, and plugins
- Infrastructure as Code: Terraform for provisioning VMs and managing multi-provider environments
- Containerization: Docker, Container Orchestration, Debugging Docker logs
- Configuration Management: Ansible and Ansible-Lint for orchestration and automation
- Monitoring: Zabbix, Graylog, Prometheus & Grafana for Cluster Monitoring
- Windows/Windows Server: Active Directory, GPO Configurations, DNS & DHCP, SCCM
- Resolved 100+ tickets related to computer networks, configurations and hardware troubleshooting
Machine Learning Intern
Campalin Innovations
July 2022 - September 2022 (3 months)
- • Developed an AI customer support chatbot using a RAG pipeline with a ChromaDB vector database for efficient knowledge retrieval, integrated with a React.js and CSS frontend, reducing average response time by ~50%.
- • Implemented a Node.js backend to manage conversation workflows, API requests, and response selection, handling simulated user queries with ~85% correct response accuracy.
- • Collaborated with mentors and teammates to optimize data preprocessing, improve model performance, and enhance UI/UX, resulting in a 30% reduction in repetitive support queries during testing.
Projects
MedSim
Virtual Patient Simulator | Jan 2025 – Present
Comprehensive medical simulation platform with AI-powered evaluation capabilities. Re-architected into microservices with Hybrid RAG pipeline on GCP.
- • Microservices architecture with React.js & FastAPI
- • Hybrid RAG pipeline using Vertex AI Vector Search & Gemini 2.5 Pro
- • Automated OSCE-style evaluation engine for real-time clinical performance analysis
- • Containerized deployment on Google Cloud Run with optimized cold-start
Clinical-Copilot
AI-Powered Clinical Assistant | Oct 2025
Advanced AI system designed to assist healthcare professionals with clinical decision-making and medical knowledge retrieval.
- • AI-powered clinical assistance and decision support
- • Medical knowledge retrieval and processing
- • Python-based implementation
Manuscript
AI Agent for Documentation | Sep 2025
An AI-agent for internal documentation and querying. Streamlines knowledge management and enables intelligent document search and retrieval.
- • AI-powered document querying and retrieval
- • Internal documentation management system
- • Intelligent search and knowledge extraction
AI-monitoring
MLOps Infrastructure | Feb 2025
Learning MLOps with Kubeflow and DevOps principles. Comprehensive monitoring and management solution for machine learning workflows in production.
- • Kubeflow-based MLOps pipeline
- • DevOps principles for ML lifecycle management
- • Production-ready monitoring and orchestration
Transgaurd
Fraud Detection AI Model | Jun 2022
An artificial intelligence model detecting if a transaction is fraudulent or not using transaction details. Machine learning-based fraud detection system.
- • AI-powered fraud detection using transaction data
- • Machine learning classification model
- • Real-time transaction analysis
Education
Master of Science - MS, Computer Science
Old Dominion University
August 2023 - May 2025
Bachelor of Technology - BTech, Computer Science
National Institute of Technology Puducherry
August 2019 - May 2023
Certifications
CKA: Certified Kubernetes Administrator
The Linux Foundation
April 2025
Networking Essentials
Professional Certification
June 2022
Introduction to Cybersecurity Tools & Cyber Attacks
Professional Certification
July 2022
Get in touch
soumith.odu@gmail.com
Connect with me professionally
GitHub
Check out my code repositories
Phone
+1 (757) 798-9998
Send a Message
Have a project in mind? Let's discuss it.