Curriculum Vitae

General Information

Full Name Rajiv Sarvepalli
Contact rajiv@sarvepalli.net
Languages English (native)

Research Areas

  • I have experience in the deep learning and cybersecurity fields through internships and research assistant positions.
  • I am interested in both Computer Vision and Natural Language Processing, especially in the intersection of the two such as Visual Q & A, image retrieval, and image captioning. I also have a broader interest in machine learning having experimented with some statistical and graphical models.
  • Generally, I have done projects in areas of computer vision, natural language processing, and machine learning and hope to continue to explore more.

Education

  • 2018 - 2021
    University of Virginia – B.S. in Computer Science
    • GPA: 3.921 (4.0 scale)
    • Major GPA: 3.976 (4.0 scale)
    • Expected graduation May 2021

Relevant Work Experience

  • Summer 2020
    Machine Learning Intern at Expedition Technology
    • Researched, assessed, and adapted state of the art object detectors using PyTorch to detect small objects in an Agile team.
    • Deployed Docker containers to AWS P2 instances through Jenkins to train networks.
    • Constructed TensorBoard live feeds to monitor validation losses and observe qualitative results quickening the model evaluation process.
    • Performed exploration of reinforcement learning libraries to find the best library with priorities of concision and readability.
  • 2019 - 2021
    Data Anyalst Intern at NetForecast Inc
    • Developed and documented data management and collection software to improve the structure and optimization
    • Designed scheduled tasks using AWS to perform constant data updates improving data quality.
    • Analyzed data from traceroutes and pings to predict router locations using support vector regression and force simulations.
    • Designed structure to follow object oriented programming principles and design patterns.

Research Experience

  • 2019 - 2021
    Undergraduate Research Assistant at University of Virginia
    • Advised by Dr. Yonghwi Kwon
    • Designed tool to analyze Docker container on Docker Hub collecting data for security analysis of Docker containers.
    • Built an Active Directory user environment through PowerShell and ESXi scripting mimicking a corporate environment.
    • Assembled information about how attacks are executed by running red-team emulators collecting usable data.
    • Preprocessed data with Python data science libraries to prepare for anomaly detection and classification tasks.
  • Summer 2017
    Research Assistant at George Mason University
    • Analyzed and organized data from IOT devices utilizing machine learning techniques to increase accuracy of models by 5\% and leveraged python machine learning libraries.
    • Implemented time series motif discovery algorithms in Python using NumPy and SciPy.
    • Created one of the first implementations of a novel matrix profile algorithm in Python from a publication.
Note: for further details on my research, please see the papers page.

Honors and Awards

Teaching Experience

  • Fall 2020
    Teaching Assistant for Computer Architecture
    • Provided weekly office hours for a Computer Architecture class with more than 300 students.
    • Taught by Prof. Charles Reiss.
    • Ran weekly laboratory sessions providing an overview and answering questions.
    • Examined instructional material for understandability and clarity through reviewing assignment's overview and instructions.

Computer skills

  • ○ Programming languages: Python, R, Prolog, C, Java, OCaml.
  • ○ Data Structures and Algorithms: Familiarity with concepts used in algorithmic competitions and machine learning research.
  • ○ Libraries: PyTorch, TensorFlow, Stable Baselines, NumPy, SciPy, Pandas, Matplotlib, Scikit-learn.
  • ○ Developer Tools: Git, Docker, AWS, VS Code, MySQL, IntelliJ.

Technical Projects

  • Check out my projects page for more information.

    Image-Caption Geolocation for Privacy
    • Used machine learning to examine the potential personal information in a social media post (image & text).
    • Leveraged name entity recognition to recognize potential information leaks within a post’s text.
    • Developed CNN for image geolocation for a small subset of locations achieving high precision and recall.
    • Made geolocation into a hierarchical classification problem through hierarchical clustering of GPS coordinates.
    • Demo Video
  • Mocking SQLAlchemy
    • Created a python library for mocking SQLAlchemy that has around 16,000 downloads and 12 GitHub stars.
    • Established modern CI/CD pipelines using GitHub Actions.
    • Documentation and source code are publicly available.
  • Docker Scraper
    • Scrapes Docker Hub for Docker images of webservers of similar types.
    • Uses variants of Docker containers to do analysis of popularity and software package version numbers.
    • Observed if people were updating their Docker containers with newer software versions to see if any vulnerabilities were left enabling data collection for security analysis of Docker containers.

Unpublished Papers

  • Anonymous Social Media Privacy
    • Senior year thesis on understanding the constraints of privacy within anonymous social media.
    • Explored the sociotechnical aspect of privacy: realizing the connection between the technical and social aspects of privacy.
    • Examined the impact of technology in changing perspective of privacy: from the individual to collective.

Relevant Coursework

  • Graduate Machine Learning
  • Machine Learning
  • Computer Vision and Language
  • Statistical Learning and Graphical Models
  • From Data to Knowledge
  • Probability

Personal Interests

  • I enjoy soccer, traveling, and video games.