Work Summaries

  • Developing an assistant that essentially serves a real-time generated powerpoint to your conversation

  • Developed an outline function, conceptual auto-correct, and auto google search questions

  • Conducted bioinformatics study to see if gene duplication of tumor suppressor genes is a evolutionarily converging feature and can be used to identify important cancer fighting genes previously unknown in the genome

  • Found ZNF560 gene which was consistently duplicated in the low cancer group compared to the high cancer group

  • Link to Github Repo

  • Participated in Kaggle competition that recreated the qualification round from Google’s 2016 annual code competition

  • Wrote a scheduler to manage drone operations to deliver products from warehouses

  • Scored 101674 with the top score being 114399

  • Link to Github Repo

  • Developed a prototype to convert audio into haptic (vibration) feedback for the deaf

  • Could feel a person’s voice

  • Built with arduino components and python

  • Link to Github Repo

  • Developed an online web-app based clipboard to easily share information between devices and people

  • Designed specifically to streamline the sharing flow using HTML5 clipboard APIs, human-readable URLs, and QR codes

  • Built with HTML, CSS, Javascript, React, Node.js, Express, Sequelize, Postgres, Mocha, Chai, Heroku, SendGrid, and Recaptcha

  • Link to Github Repo

  • Analyzed mock pandemic data to determine a scoring function to correlate specific characteristics about each city/county/state and it’s associated spread of infection providing insight on areas of susceptibility, effectiveness of interventions, and etc

  • Contracted to help manage and develop an IOT solution to improve transparency, provide insights, and optimize efficiency for RMS (Rail Management Services) one of America’s largest intermodal shipping companies

  • Generally led the execution side of the project

  • Assisted with all facets of development: product design, frontend development, and backend development

  • Helped hire, vet, and manage contractors while coordinating closely with the project manager

  • Developed in collaboration with Gnar.io

  • Contributed to development of software as a service to make using computer vision as easy as setting up an email filter

  • Assisted with multiple facets of development such as product design, frontend development, backend development, API design, and various product demos

  • Product design work included initial conceptualization and collaboration with an external design team

  • Frontend development included new feature development, bug fixes, Github issue management, and collaboration with an external development team

  • Backend development included new feature development, bug fixes, API design, and collaboration with the internal backend development team

  • Product demos covered a wide range such as a counterfeit currency detector, recycling assistant, data fusion demos, a data explorer to comb through various events and camera feeds, and a sophisticated google voice assistant with advanced navigation and querying capabilities

  • A twitter/reddit hybrid for fun, practice, and learning about building online communities

  • Built with HTML, CSS, Javascript, React, Redux, Node.js, Express, Mongoose, Mongodb, and Heroku

  • Link to Github Repo

  • Contracted to create a fully automated fluorescent multi-spectral microscope prototype with sub-micron resolution, a control system (covering XYZ axis motion control, LED brightness and selection, collimator adjustment, polarizer toggle, and polarizer rotation), automated giga-pixel image capture, and a display/control interface

  • Managed all planning and engineering from idea to working prototype covering all hardware, electronic, software, and optical aspects

  • Contracted to enable credit card swipe payments for for Purdue Alumni Association to replace hand entered data

  • Developed a credit card Chrome extension that enables swipe payments by collecting raw data from a USB connected card swiper, parsing the data, and pasting it into the web form

  • Self-contained extension written with HTML, CSS, and vanilla javascript with no outside dependencies, various attack surfaces covered, completely transient handling of credit card information, and requiring minimal permissions

  • Developed in collaboration with Gnar.io, which provided the visual design

  • Developed a custom journaling application enabling users to easily view and organize their memories

    • Giant Timeline View of your Life

    • Photos, Gifs, Videos, and Video previews

    • Add new entries instantly anywhere

    • Entries don't require a full date

    • Dropbox backup

    • Companion AI

  • Enables users to define a journal entry date and time with whatever you know about when it happened (such as "2016 May", "2016 May 23rd", "2016 May 25th 5pm", "before 2016", “after 2016 November”). And it still sorts with whatever available information using a custom recursive sorting algorithm, so if you try to put something somewhere that doesn’t make sense, it stops you from doing that and keeps your entries chronological order consistent

  • Includes custom frontend components such as variable component size lazy loader and click to journal entry scrollbar

  • Integrated with Dropbox Api along with customized external libraries to provide local syncing and full media support

  • Includes a backend companion AI to help you write your journal entries using custom heuristics with a combination of summarization and sentiment analysis to identify areas to inquire about

  • Built with HTML, CSS, Javascript, React, Flux, Python, Flask, Google Natural Language API, and Heroku

  • Link to Web App | Link to Github Repo

  • Helping save the rainforest by improving satellite image classification. Images can be classified as some combination of forest cover, roads, cloudy, habitation, mining, slash and burn, and others

  • Implemented a fine-tuned deep convolutional neural network based solution with custom features. The network architecture fed the image data into the pre-trained Xception neural network as the initial layers, and then the network output was combined with image RGB statistics into a final output layer

  • Achieved accuracy score of 0.90 with the winning accuracy of 0.93 as reference

  • Utilized AWS EC2 GPU instances with customized images to accelerate development

  • Used Bash, Python, Scikit-learn, Numpy, Pandas, Tensorflow, and Keras

  • Developed companion web app to help understand and fine tune canny edge, hough line, and harris corner detection algorithm settings. Built with HTML, CSS, Javascript, React, Python, Flask, openCV, and Heroku

  • Link to Github Repo

  • Contributed to development of a nanopore-based DNA sequencer that enables long continuous DNA reads vital to analyzing the human genome

  • Conducted key research and provided key guidance to enable accomplishment of a product goal (minimum 10X multiplier to sequencing yield via improvements to hardware lifespan) a year ahead of schedule using a combination of system engineering and failure analysis requiring interdisciplinary knowledge and collaboration in chemistry, physics, molecular biology, engineering, and experimental design

  • In addition, the increased throughput early in development will potentially save the company millions of dollars in R&D cost over time

  • Developed a 95% accurate, robust, scalable, and adjustable signal processing algorithm that identifies and quantifies a key signal pattern on gigabytes of data within minutes using a variety of statistical methods. Used Python, Numpy, Scipy, Pandas, and Scikit-learn

  • Assigned patents: Patent1 Patent2

  • Invented a powerful portable smartphone microscope to combine the features and portability of a smartphone with the ability to see into an invisible world

    • Simple versatility with compatibility with nearly any smartphone or tablet with auto-focus+exposure

    • Flexible power with 30X, 170X, or 340X Optical magnification with ~5 micron resolution

    • Ultraportable enough to fit in your pocket, it goes where you go. Runs for hours on two AAA batteries

  • Successfully crowdfunded over $30,000 which is among the 95% percentile of campaigns Link to Kickstarter | Link to Indiegogo

  • Communicated with press resulting in articles in TechRepublic, Fastcompany, and L'Atelier BNP Paribas along with handling public relations among backers

  • Manufactured the microscopes in-house through a combination of advanced technologies such as 3D printing, resin casting, laser-cutting, waterjet cutting, machining, and etc. Optimized and failure analyzed the manufacturing process

  • Developed a low cost (less than $200) handsfree computer interface using an eyetracker and EMG (electrical muscle impulse sensor) combo to contribute to the general wealth of knowledge

  • Used python to interconnect an external library that provided a mouse API, the eyetracker SDK, and the arduino that’s connected to the EMG

  • Link to Instructable | Link to Github Repo

  • Ran a tutoring service that taught math, college-level chemistry, physics, organic chemistry, and biology

  • Helped students move up from Ds and Cs to As

  • Contributing factor to me achieving a 98.5 percentile score on the MCAT

  • A song I wrote that marked the beginning of a journey