Computer Science Master's + Bachelor's graduate from the University of Southern California
Working at USC Multi-omics Mass Spectrometry Core
mistrik[at]usc.edu | mistrikdenis[at]gmail.com
View My LinkedIn Profile
View My Resume
A full-stack comprehensive graphing, visualization, and data analysis tool for SEC corporate filings. Includes the ability to create your own custom dashboard of SEC data, from all available years, and have up to 5 companies compared at one time.
Additionally you can perform aggregate data science with PCA dimension reduction and HDBSCAN clustering.
Runs on its own dedicated VPS with a custom SQL database I populated from raw SEC EDGAR data.
Available at [filingsvisualized.com]
This python library utilizes tkinter to create a GUI and a canvas to visualize how a simple genetic algorithm would traverse a specific maze. Comes with various customizable variables, such as mutation rate, generation size, sequence length etc. Once the algorithm successfully reaches its goal node, the user will be prompted with the ability to graph various statistics relating to the genetic algorithm.
This game features the standard midlane setup found in games such as Dota 2 or League of Legends. The player has 4 unique abilities to choose from and has to face an AI that uses advanced heuristics which are weighted by ChatGPT API calls and its game state analysis

This game features two lanes with two towers and a base, the player is given a deck of poker themed cards that they can place on the battlefield that will then battle it out with the enemy
Implemented custom flow field logic for navigation to ensure performance and consistency.
Integrated a ChatGPT AI enemy, which is prompted with the game state and will decide what units to place if it has the resources for it
Android application replicating the functionality of a food delivery app such as DoorDash or Uber eats. It can track a persons profile, their daily caffeine intake, give recommendations, and can give an expected delivery time in addition to visually showing the route taken. Created using Android Studio (Java) with a Firebase backend and the MapBox API for the mapping/directions.

Web based version of pictionary, where the players can all join a singular chat room with one person acting as the drawer and rest guessers.
Implemented using Apache TomCat, basic html frontend with p5.js integration, and a java backend