PrathameshUravane
MS in Applied ML at University of Maryland (GPA 4.0). I build production-ready AI systems — from data pipelines and LLM agents to computer vision and clinical AI — that ship to real users and create measurable impact.
AI agents · RAG pipelines · Production ML systems for real-world clients
Who I Am
About Me
I'm an AI engineer who translates complex requirements into clean, measurable systems. My work sits at the intersection of machine learning engineering, computer vision, and health informatics — building pipelines that go all the way from raw data to production APIs.
What sets me apart is the full-stack perspective: I can design a model architecture, wire it into a FastAPI backend, containerize it with Docker, deploy it to AWS, and ship a polished Next.js dashboard on top — independently, from brief to delivery.
My research roots keep me honest about what matters: rigorous evaluation, clear metrics, and solutions that hold up under real-world conditions. Whether it's a RAG agent evaluated on 150 business test cases or a CNN achieving 96% on MRI classification, I measure everything.
Outside of work, I officiate 11v11 soccer — which has made me very good at making quick, accurate decisions under pressure. Turns out that skill translates well to ML engineering too.
Production-First
I ship systems that run in production — not notebooks. FastAPI backends, Docker containers, AWS deployments.
Global Experience
Worked across India, Singapore (NTU), and Peru (UMA). Cross-cultural collaboration is second nature.
Published Researcher
5 peer-reviewed papers in IEEE/Elsevier spanning healthcare AI, computer vision, and predictive analytics.
Based in DC/Maryland
MS student at University of Maryland, College Park. Actively seeking Summer 2026 internships.
Career Journey
Where I've Worked
Intramural Sports Official
CurrentUniversity of Maryland
College Park, MD
- Officiate 11v11 intramural soccer matches, enforcing rules and maintaining match control in high-pressure environments.
AI Engineer
UMA – Universidad María Auxiliadora
Lima, Peru
- Designed and deployed an AI-powered virtual lab simulator with an NLP feedback pipeline, increasing student engagement by 34%.
- Built a real-time student attentiveness monitoring agent using computer vision (face detection, facial landmarks), reducing manual review time by ~70%.
- Collaborated with cross-functional educators and developers; documented AI solutions for ongoing client support.
Student Researcher Intern
Energy Research Institute @ NTU
Singapore
- Implemented a GAN-based model to synthesize realistic road scenarios from pedestrian data, automating training-data generation for autonomous vehicle perception systems.
- India Connect Research Fellowship — competitive national program for research exchange.
AI Engineer Intern
Yodda Elder Care Technologies
Pune, India
- Developed an OpenCV-based fall-detection agent with 95% accuracy using pose estimation and classification models.
- Integrated automated alarm triggering and snapshot delivery in milliseconds, eliminating manual monitoring for elderly care facilities.
- Delivered fully documented Python backend with integration guides enabling independent client extension.
Student Researcher
VU Research Centre — Health Informatics
Pune, India
- Trained and evaluated a CNN model for brain tumor classification from MRI images, achieving 96% accuracy.
- Produced model pipeline documentation and performance metrics for the research team.
What I've Built
Featured Projects
Portfolio links coming soon — drop me a message for demos, code walkthroughs, or live previews.
InsureLLM
RAG-Powered AI Agent for Business Knowledge Automation
Built an end-to-end RAG AI agent for an InsureTech company to automate employee access to policy, contract, and HR knowledge. Replaces manual lookups with a conversational, LLM-powered workflow.
- MRR of 0.875 on 150 business test cases
- 95.7% keyword coverage — production-ready accuracy
- Semantic search + re-ranking pipeline
MediNotes AI
Generative AI Integrated into Clinical Business Workflow
Production SaaS product that automates clinical documentation. Doctors input consultation notes and receive AI-generated visit summaries, action items, and patient-facing emails via real-time streaming.
- Full-cycle SaaS: auth, billing, multi-tenancy
- JWT auth + Clerk Billing for subscription-gated access
- Dockerized, deployed serverlessly on Vercel + AWS
Vehicle Predictive Maintenance
End-to-end predictive analytics pipeline for vehicle failure forecasting with automated model retraining on data drift — zero manual intervention required.
Student Attentiveness Monitor
Multi-threaded computer vision pipeline using facial recognition and landmark tracking to assess student attentiveness in real-time during remote classes.
Brain Tumor CNN Classifier
CNN model for brain tumor classification from MRI images, published as part of health informatics research at Vishwakarma University's research centre.
Autonomous Vehicle GAN
GAN-based model generating realistic road scenarios from pedestrian data to expand AV training datasets with synthetic but realistic traffic environments.
Full portfolio with live demos
Coming Soon
In-depth case studies, GitHub repos, and live demos will be linked here as they ship.
github.com/uprathamTechnical Arsenal
Skills & Stack
✓ AI Agents & Automation
✓ Machine Learning
✓ Computer Vision
✓ Backend & MLOps
✓ LLMs & GenAI
✓ Frontend & Data
Full Tool Belt
Academic Foundation
Education & Research
Master of Science in Applied Machine Learning
University of Maryland, College Park
College Park, Maryland
- Perfect GPA in first graduate semester
- Focus areas: deep learning, NLP, computer vision, MLOps
India Connect Research Fellowship
Nanyang Technological University Singapore
Singapore
- Competitive national fellowship for research exchange
- Research: Gen AI for Autonomous Vehicles, GAN-based road scenario synthesis
B.Tech in Artificial Intelligence & Data Science
Vishwakarma University
Pune, India
- AI/ML Core Team Member — Google Developer Student Club
- Multiple IEEE/Elsevier research publications during undergrad
Publications (5 Papers)
CNN-Based Brain Tumor Classification from MRI Images
ANN-Based Dyslexia Detection with Gamified Data Collection
Student Dropout Prediction using Pre-trained ML Models
GAN-Based Synthetic Road Scenario Generation for AV Perception
Real-time Student Attentiveness Monitoring via Computer Vision
Certifications
Contact
Let's build something that is useful,measurable, and shipped.
I am open to internships, product engineering roles, and research collaborations where AI needs to move from prototype to production.
Prathamesh Uravane · AI Engineer · ML Researcher
Designed for clear, measurable, and recruiter-friendly storytelling.