I work as an Optimization Engineer solving problems for the brewing communications revolution at Kythera Space Solutions. I graduated from Rensselaer Polytechnic Institute where I studied data analytics, cryptography, and software development.
I have a diverse background with an undergraduate degree in history. I have work experience in a variety of professional environments including medical records, commercial management, and the manufacturing industry. I like to have fun and play games and listen to music.
Contact Me
Space Pyrates
Space Pyrates is a rogue-like PC application in which the user plays as a pirate in a futuristic science-fiction evironment. Besides coding mechanics the development of Space Pyrates involved a significant amount of animation and graphic design. Both GameMaker and Unity were used as a scripting IDEs and Adobe Creative Suite was employed for graphic design and animation.
I swore to myself when I was small that when I was old and rich I would have all the armies I couldn't afford when I was young... that day has come. I realized, however, that the rules of warhammer had slowly warped over time to be less like DnD and more like Magic which has its place to some extent. To truly give my pieces the complex beautiful world they deserved I made a hybrid ruleset of all the coolest mechanics. I then began playtesting the ruleset which also yields a living lore for my miniatures.
I am writing my own customized Linux kernel while contributing to coursework at Rensselaer Polytechnic Institute. It is technically a Linux 2 kernel but it is completely compatible to boot Debian or Ubuntu. This is project has relatively extensive C programming from customized SysCalls and Interrupts to a re-interpreting TCP socket layering and IP data structures.
Language used:
C
Project type:
Networking, Operating Systems Development
Cryptosystem
I have constructed a primitive cryptosystem using mathematic principles of cryptography rather than simply making calls to existing libraries. It has the functionality to effectively encrypt and send files between users. It also has the functionality to check if users have the same files withouth betraying information about non-shared files. The necessary coding was done in Microsoft Visual Studio, eclipse and IntelliJ.
When I was getting my graduate degree I focused basically on taking the furthest out hardest classes to make one master of the computer, ,humanity's most complicated and powerful tool like a role-playing character seeking dungeons to level up in. This lead me of course to study both data analytics and cyber-security which are not directly related besides being computer classes with a lot of pre-requisites. Now, I have taken up a position as an Optimization Engineer in a field where you basically can't take any classes because it is unsolved even by academics: Space Solutions. At Kythera I solve the real world problem of providing satellite communications to the earth. Developing algorithms and writing software to perform them has lead me to work with a wide smattering of the largest satellite manufacturers, satellite operators, and content providers. Being on the bleeding edge of a burgeoning industry is already as exciting as it sounds but our industry is space... its like a little kid dream of "Waddya wanna be when you grow up?"
Languages used:
C++ for all the heavy lifting with Python, R and javascript on the side for littler things.
Project Type:
Generating a toolset to be the generic Operating System for space solutions.
While working at Booz Allen Hamilton I led a group of interns in a project integrating data analytics and cybersecurity. ADS-B, the protocol used to communicate gps positioning of airplanes is completely unencryped and unauthenticated which presents a myriad of security risks to modern aviation generally. While this threat is both dire and immediate I endeavored to develop an economically feasible solution by training an Artificial Neural Network to classify ADS-B messages effictively using machine learning instead of cryptography to establish secure communication with a software update rather than re-writing the entire protocol. The product, FlightSecure consists of a full web-app which visualizes planes on a map which is connected to the classifier which can filter out attacks and false messages from being displayed to pilots, air-traffic-control, or any other potential users. This project was completed using the Tensorflow and Keras libraries and both R, and Python in the Jupyter Labs IDE.
I was selected by Microsoft to lead a group of my peers for a semester in a software development project. We were tasked by the Sway team to present a prototype for creating and improving a web site's First-Time User Experience. Microsoft did not want to end up using jQuery or libraries which implement jQuery a ubiquitous but antiquated library for a process that while invaluably important should require only one repetition per user. We solved the problem by writing a new Javascript library with no added libraries with the abilities to customize tutorials as well as containing lightweight defaults which could immediately expedite this aspect of web development.
Language used:
Javascript, JSON, HTML/CSS
Project type:
Software Development, Web Science, User Experience
Project Link:
The deliverable resulting from this project is owned by Microsoft so I cannot link to it although they may make it public themselves..
Presenting Technical Information to a Novice Audience
I have demonstrated ability to communicate complex concepts to stakeholders who may not be technically trained while working as a Teaching Assistant as Rensselaer Polytechnic Institute. I have been selected to do so three times in multiple departments designed lectures syllabi and giving frequent presentations concerning concepts of Data Science, Xinformatics, and Cybersecurity. I have also recorded educational videos to complement course material.
Automating Foreign Currency Trading With A System of Neural Nets
My most ambitious project yet is a set of three neural net models piped together into an automation system to trade foreign currencies. It will start with a prediction module using LSTM which is often used in image recognition but can be repurposed to predict time-series price data. The predicted prices the LSTM outputs will be acted upon by a generative adversarial network (GAN) to create trade data from the predicted prices. These trades will then be fed into a machine learning classifier which will judge them for profitability and therefore train the generator to make more effective trades. Once fully trained the system will be connected to an automated training system written with the one of the APIs on Forex or OANDA. The neural networks will be written in Tensorflow and trained using AWS instances.
Concatenated data on the changing price history of a variety of online cryptocurrencies including: BitCoin, Ethereum, and Monero as well as many others. The purpose of this collection was to discover any relational differentiation between brands of cryptocurrency with respect to trends that appear to affect the market as a whole. Data was collected by gaining access the Kaggle kernel. Collected data stored in CSV files was then organized and moved into XLSX format and then on into relational database. I have uploaded a simple R script which demonstrates how it can be useful to keep the data in a relational database rather than separate CSV files. Also provided are some pdf images of visualizations made by running the R script on the database.
Language used:
SQL, R
Project type:
Financial Analysis & Data Collection
This is a tool for collecting data associating the box office gross of popular movies with their feedback on review and ratings websites such as Rotten Tomatoes and Metacritic and box office data was obtained from Box Office Mojo. The purpose of this collection is to analyze the strength of the relationship between a movie's critical acclaim and its box office gross. The data is acquired using the daily python script which creates a JSON file of aggregated data. I have posted an example of its output.
A tool created for scraping data on Reddit.com allowing the user to choose sub-reddits to crawl through and keywords to search for. Data is acquired by accessing the Reddit API utilizing python and the python reddit api wrapper, or PRAW. The data is stored as a collection of JSON files. This is a reusable tool that can be used for various data collections, I have included an example of its output.
Language used:
Python, JSON, Reddit API
Project Type:
Data Collection & Software Development
