Benjamin Wang

Experience

• 

  University of North Texas

  Aug 2019 - May 2021

  Undergraduate Student Researcher

  • Researched optimal vector parameters to maximize sparsity for a custom neural network chip by analyzing the hidden layers in convolutional neural networks (CNN).• Collected and graphed data such as MobilenetV2, VGG19, and more with Python libraries Hoplite and plot.ly which provided layers with optimal sparsity.• Developed a tool called See-NN in order to visualize a CNNs hidden layers which allows for researchers to further research misidentification due to bit-wise errors, differing training methods for varying CNNs, and error injection attacks Show less

• 

  Huntington Learning Center

  Aug 2021 - Nov 2021

  K-12 Tutor

  - Developed custom curriculum for students which allows for personal pacing and growth.- Adjusted teaching methods on the fly depending on how successfully it was implemented by the student.- Bonded with students through personal anecdotes and common interests.

• 

  Capital One

  Jun 2022 - Aug 2022

  Technology Intern

  • Modernized the adverse action infrastructure by hosting a full stack app on AWS which replaces over 1000 lines of SQL.• Pulled up-to-date information on an auto loan applicant using JDBC, JPA, and AWS connections allowing the processteam to more accurately process the millions of adverse action letters daily.• Worked with other interns to safely store and retrieve sensitive information by integrating multi-factor authenticationand Jenkins configurations

• 

  ACM UTD

  Dec 2022 - Dec 2023

  Vice President

  ACM @ UTD strives to redefine the landscape of computer science through collaboration, outreach, and inclusivity. We help bring the worlds of academia, industry, and more to the forefront by empowering our officers to push the limits of education.

• 

  Capital One

  Jun 2023 - Aug 2023

  Technology Intern

  •Constructed a knowledge graph in Stardog consolidating over 10 billion rows of data over 8 production databases, centralizing exploratory data analysis for machine learning and data correction across Bank, Account, and Data Science teams.• Developed a dashboard that uses Stardog’s HTTP API to query over 10 billion rows of live data and for any transaction flagged by external fraud detection models it will first retrieve the bank account associated with the transaction and then aggregate the account owner’s contact information.• Quantified Stardog Explorer’s scalability by employing Cypress-like functions that measured SPARQL query runtimes relative to the ontology’s size; demonstrating how proper ontological subsets speeds up data exploration and visualization by a factor of 20x. Show less

• 

  Capital One

  Feb 2024 - now

  Associate Software Engineer