Sumedha Vadlamani

About Me

Sumedha Vadlamani

I'm a recent Data Science graduate from the University of Maryland, College Park, with a strong foundation in AI, computer vision, and multimodal systems. My work bridges the gap between research and practical applications across areas like generative AI, NLP, and biomedical imaging.

At UMD, I worked on several cutting-edge projects—training GPT-2 variants with different attention mechanisms, fine-tuning Mistral 7B for poetic device classification, and building a Retrieval-Augmented Generation (RAG) system using MongoDB and transformer models. These experiences sharpened my skills in large model training, efficient deployment, and interpretability. Previously, I earned my B.Tech in Computer Science with a specialization in IoT from VIT Vellore, where I received merit scholarships and published research on liver tumor segmentation and Hepatitis C prognosis using deep learning.

I've interned at Causify AI , optimizing Google Drive APIs and building scalable ETL pipelines. Earlier, I was a research assistant at VIT University, where I developed deep learning pipelines for medical imaging tasks. I also enjoy building and deploying ML systems end-to-end using tools like Docker, Flask, Azure, and SQL-based backends.

Outside of work, I'm passionate about reading articles on emerging AI technologies and exploring technical blogs on Medium. In my free time, I enjoy doing vector art illustrations, going on hikes, and food hopping to discover new cuisines.

I'm currently looking for full-time roles in AI, data science, or machine learning starting June 2025. If my work aligns with your team's goals, I'd love to connect!

Skills

    Programming and ML
  • Python
  • SQL
  • R
  • Sci-kit Learn
  • PyTorch
  • TensorFlow
  • OpenCV
    Tools and Platforms
  • Databricks
  • Docker
  • Streamlit
  • Power BI
  • Microsoft Office
  • MongoDB Atlas
  • Git
    Cloud and Databases
  • Microsoft Azure
  • Apache Spark
  • PostgreSQL
  • Google Cloud Platform
  • AWS

Professional Background

Education

University of Maryland, College Park

Aug 2023 - May 2025 | College Park, MD

M.S in Data Science

  • GPA: 3.88/4.0
  • Coursework:
    • Machine Learning
    • Multimodal Foundation Models
    • Probability and Statistics
    • Deep learning
    • Data Representation and Modeling

Vellore Institute of Technology (VIT) Vellore

July 2019 - June 2023 | Vellore, India

B.Tech | Major: Computer Science Engineering | Minor: Internet of Things

  • GPA: 8.89/10
  • Coursework:
    • Data Structures and Algorithms
    • Computer Architecture
    • Database Management Systems
    • Operating Systems
    • Advanced Calculus
    • Micro-processors and Micro-controllers
    • Privacy and Security in Internet of Things
    • Applications of Differential Equations and derivatives
  • Research: Machine Learning Research Assistant at VIT University. Contributed to two published works—"Liver Tumor Segmentation Using Deep Learning Neural Networks" (book chapter) and "Hepatitis C Severity Prognosis Based on Blood Biomarkers: A Machine Learning Approach" (research paper).
  • Awards: Received a merit scholarship for outstanding academic performance (2021-2022) and the Raman Research Award for excellence in undergraduate research.
  • Extra-curricular activities: Core Committee Member of IEEE-Professional Communication Society at VIT University. Helped organize technical events, managed a team of 50 members, and promoted wide participation. Also served as a Core Committee Member of the LEO Club (NGO), organizing donation campaigns, visiting orphanages, and leading awareness initiatives.

Research and Work Experience

Causify.AI Inc

Feb 2025 - March 2025 | Remote

Data Engineer Intern

  • Extracted and processed structured data via custom web scraping workflows.
  • Developed robust unit tests to ensure system reliability and maintainability.
  • Refactored legacy codebases to improve performance and readability.

ML Research Assistant Intern | VIT University

Aug 2021 - June 2023 | Vellore, India

Guide: Dr. Sumit Kumar Jindal

  • Publication: "Liver Tumor Segmentation Using Deep Learning Neural Networks" - Book Chapter in Current Applications of Deep Learning in Cancer Diagnostics, Taylor & Francis, 2022.
  • Developed a hybrid U-Net deep learning model for liver tumor segmentation, achieving 75% accuracy on clinical imaging datasets.
  • Performed data augmentation and preprocessing to improve model generalizability and training stability.
  • Publication: "Prognosis of Hepatitis C Severity Based on Blood Biomarkers Using Supervised Machine Learning Algorithms" - Journal of Electrical Engineering & Technology, Springer, 2022.
  • Designed machine learning models to predict Hepatitis C severity, achieving 98% accuracy using Gradient Boosting.
  • Built and optimized preprocessing pipelines with advanced feature engineering and hyperparameter tuning techniques.

Project Portfolio

GPT2-Stripped-Comparative-Insights
GPT2-Stripped-Comparative-Insights

Implementation of a simplified GPT-2 model and experimenting with various attention mechanisms and positional embeddings

Poetic-device-classifier
Poetic-device classifier

Fine-tuned the Mistral 7B language model to classify poetic devices in custom text datasets, improving accuracy through LoRA adapters and quantization techniques

Human Activity Recognition with wearable sensors
Human Activity Recognition with wearable sensors

Classifies human activities using data from wearable sensors like accelerometers and gyroscopes

Flight-Delay-Analysis
Flight-Delay-Analysis

API-driven data analyses on some of the busiest airports globally, extracting key insights to inform the training of state-of-the-art machine learning models

Context-Aware Document Retrieval and Generation System with RAG and MongoDB
Context-Aware Document Retrieval and Generation System with RAG and MongoDB

Built a RAG system using MongoDB and Mistral 7B to retrieve and generate answers from PDF documents with semantic search

Next Generation Parking System using OpenCV
Next Generation Parking System using OpenCV

Developed a real-time smart parking system using CNN and OpenCV to detect available parking slots from CCTV feeds, with live updates via Flask and digital booking through a mobile app

Farm2Kitchen AI
Farm2Kitchen AI

Developed a Flutter-based smart kitchen app integrated with Firebase for real-time inventory, recipe suggestions, and farm-fresh produce tracking, enhancing food planning and sustainability

Data-Driven Sales Analysis Using SQLAlchemy ORM and ML
Data-Driven Sales Analysis Using SQLAlchemy ORM and ML

Analyzed supermarket sales data using SQLAlchemy and visualized insights with Matplotlib; deployed a Flask-based sales prediction API using Docker.

Blogs

blog9
Multimodality and Gen AI- paper reviews

blog8
Hepatitis C Severity Prognosis: A Machine Learning Approach

Developed machine learning models to predict Hepatitis C severity based on blood biomarkers, achieving 98% accuracy using Gradient Boosting and extensive feature engineering

blog7
Liver Tumor Segmentation Using Deep Learning Neural Networks

Worked on liver tumor segmentation using deep learning; deployed and fine-tuned a hybrid U-Net model, achieving 75% accuracy through extensive data preprocessing and comparative analysis

Awards

Vikram Setty

Merit Scolarship

April 2022

Awarded merit scholarship for outstanding academic performance in the computer science engineering with specialization in IoT cohort during the year 2021-22.


Vikram Setty

Raman Research Award

June 2023

Awarded the prestigious Raman Research Award Awarded the prestigious Raman Research Award in recognition of my contributions to research and publication during my undergraduate studies at VIT. The award honors impactful scholarly work published in reputed journals and encouraging a culture of academic excellence.

Hobbies

Sumedha Vadlamani

I love spending my free time engaging in a mix of creative and meaningful activities. One of the things I care deeply about is community involvement, and I actively participate in blood donation drives. I find it incredibly rewarding to be able to contribute in such a direct, impactful way — it's a small act with the potential to save lives, and I try to donate regularly whenever drives are organized near me.

Creatively, I enjoy making vector art illustrations, often drawing inspiration from pop culture, music, and people I admire. It's a therapeutic process that lets me unwind and express ideas visually — some of my favorite pieces have come from just experimenting with new color palettes or styles.

I also consider myself a bit of a food enthusiast. I love exploring different cuisines, food hopping around cities, and trying dishes that are unique or offbeat. Occasionally, I try to recreate those meals at home, experimenting with recipes and ingredients along the way.

While I enjoy the outdoors, my interest in hiking is more casual — I take the occasional trail when I get the chance, especially if there's a great view or local spot to discover at the end.

Overall, I'm drawn to activities that combine creativity, exploration, and social contribution — whether that's through art, food, or simply giving back where I can.


Contact Information

Location:

San Diego, CA 92129