Hi! I'm Surya.

I'm a driven AI and biology student with 3 years of experience turning industry problems into data-driven solutions.

Experience

As a dual major in Computer Science and Biology, I've accumulated a wide range of experiences in both fields which I ultimately hope to synthesize together one day.

AI Engineering Intern at Thermo Fisher Scientific | May 2021 - Present

  • Created a training dataset from limited cell plate images, developed a custom U-Net computer vision model to identify viable stem cells, and optimizing the model to perform on a microscope in less than 3 seconds.
  • Deploying stem cell identification model to two edge devices, Nvidia Jetson Nano and Intel Tiger Lake, using an AWS pipeline including S3 bucket, Sagemaker Neo, Edge Manager, and MLFlow to analyze metrics.

    • AI Researcher at Space ML | July 2020 - Feb 2021

  • Built, optimized, and integrated a meteor classifier LSTM model trained on night sky light sources, a key part of the NASA-funded CAMS project, which aims to identify unknown meteors and predict their parent bodies.
  • Developed a unique data augmentation technique which increased precision and recall to 95% from the original 65%. This work was the basis for my paper presented at EGU 2021, an international AI conference.

    • Open Source AI Researcher with mentors Anirudh Koul and Siddha Ganju | May 2020 - July 2020

  • Constructed a CNN-LSTM pedestrian trajectory prediction model trained on a pedestrian video dataset with time-series accelerometer annotation. Aiming to assist blind pedestrians, this model was developed to show how AI can be used for social good, and preliminary results were presented at CVPR 2020, a top AI conference.

    • Developer at MIT GrandHack DC | August 2019

  • Developed Breathe Easy, Designed a remote asthma monitoring device inspired by continuous glucose tracking in diabetes patients which could eliminate up to 60% of preventable ER visits for asthma patients by detecting anomalous airflow patterns.
  • Won Runner-up in the Access to Healthcare track, leveraging domain knowledge from pulmonologists on my team.

    • Research Intern at Radify Labs | May 2019 - July 2019

  • Built a pipeline for heart disease patient data with Numpy and Pandas, using Tableau to visualize trends in patient data.
  • Constructed random forest and decision tree diagnostic models using Tensorflow and Keras. These diagnostic models were built from the UCI Cleveland heart disease dataset, and were able to reach 85% diagnosis accuracy.

    • Education & Skills

    I'm currently a fourth year student and Echols Scholar at the University of Virginia with a 3.9 GPA, planning to graduate in May 2022.

    AI / ML

  • Tensorflow
  • Keras
  • SciKitLearn
  • Pytorch
  • Weights and Biases
  • Jupyter Lab
  • AWS

    • ML Domains

  • Computer Vision
  • Clustering and Classification
  • Statistical Modeling
  • Quantitative Analysis
  • NLP

    • Programming Languages

  • Python
  • Java
  • C
  • C++
  • SQL
  • R
  • Lean
  • Git

    • Data Science & Visualization

  • R
  • NumPy
  • Pandas
  • Excel
  • Tableau
  • Matplotlib

    • Biological

  • PCR
  • Western / Southern Blotting
  • Mammalian Cell Culture

    • Web Development

  • HTML
  • JavaScript
  • Django
  • Heroku
  • Travis CI

    • Relevant Coursework (Biology)

  • Biostatistics
  • Genetics & Molecular Biology
  • Regenerative Biology
  • Cell Biology

    • Relevant Coursework (CS and Statistics)

  • Artificial Intelligence
  • Databases
  • Machine Learning
  • Algorithms
  • Vision and Language
  • Embedded Computing & Robotics
  • Advanced Software Development
  • Statistical Analysis with Python

    • Awards & Memberships

    UVA Deloitte Case Competition 2020

    Winner of the 2020 UVA Deloitte Case Competition and moved onto the national competition at Deloitte University in Fort Worth, Texas.

    MIT Grand Hack DC 2019

    Runner-up of the 2019 MIT Grand Hack DC on the 'Access to Healthcare' track for 'Breathe Easy', a personalized remote asthma symptom monitoring device.

    TEDxUVA Competition 2019

    Selected to speak at the 2019 TEDxUVA Student Speaker Competition.

    Memberships

  • Writer and Developer for UVA Machine Learning Club
  • Member of the Jefferson Literary & Debating Society, one of the oldest debate societies in North America
  • Member of UVA Google Student Developer club
  • Member of UVA Moot Court, a constitutional law and debating team

    • Certifications & Publications

    AWS Certified Cloud Practicioner

    Certification Date: June 2021; Expiration Date: June 2024

    EGU 2021

    Primary author of “It’s a Bird, It’s a Plane, It’s a Meteor!" presented at EGU 2021 on April 28, 2021

    CVPR 2020

    Work presented at Computer Vision and Pattern Recognition 2020, a top-tier global AI conference. This project, a pedestrian trajectory predictor, was done with mentorship from Anirudh Koul and Siddha Ganju.

    Get in touch

    Don't hesitate to contact me with any questions or comments!