Grant Mak Email & Phone Number

Cary, North Carolina, United States

• Built full stack server side applications for both internal and customer facing services. The work I did drioes internal device development and increased user experience for Garmin customers.
• Overhauled the front and back end of Garmin's support logging service to make it easier to configure and receive feedback from external devices.
• Wrote a new recommended apps algorithm for the Garmin ConnectIQ store to provide Garmin customers more personalized recommendations.
• Designed and implemented a system to passively collect, aggregate, and display Key Performance Indicators for use by internal device developers and external app developers.

Designed a multilayer, backpropogated feedforward neural network. The network was trained on a set of medical data, with the network being able to predict post discharge mortality rate with an accuracy of over 97% using both socioeconomic factors and medical conditions as inputs.

Was a member of an intensive, three-semester long design program aimed at developing industry skills such as design, root cause analysis, and Agile workflow. Designed a light box that blinks at different frequencies using Eagle and Fusion360. Designed a low cost dual IMT and EMT respiratory device that could train with a high resolution over a large.

Gave feedback on Arduino code written to integrate sensors on a Parallax BOE-Bot to collect data, follow a path, and send and receive signals to and from other BOE-Bots. Taught basic breadboard circuitry and instrumentation.

Designed and implemented a 5-stage pipelined processer in Verilog with a MIPS-like instruction set to create a karaoke based light and music show on a Christmas tree using basic arithmetic and memory operations.

Cary, North Carolina

• Collaborated with Duke University to lay the groundworks for a relationship with Garmin to use Garmin devices to drive biomedical teaching and research.
• Wrote a Python program that uses OAuth1 with HMAC-SHA1 protocol and REST methods to pull patient data off of Garmin Connected Services into a MongoDB database.
• Implemented a proxy server that takes intermediate HTTP requests from the app and forwards the full HTTP request to Garmin Connected Services.
• Designed a frontend for the application using QT, and the application was distributed as a Docker container with another linked container containing the MongoDB database.

Jacksonville, Florida

Collected and analyzed raw hyperspectral data in MATLAB to obtain information on wavelength spectra, chromatic content, and chromatic contrast between an object and its surroundings. Wrote a custom MATLAB script that implements a logical mask to automatically zoom in on an object of interest.