I'm a motivated Computer Science student looking to leverage my skills learned through previous internship experiences as they apply to impactful future opportunities.
Worked on the Ads Product Engineering team to increase inventory by showing ads in Topic Feed. Also designed aggregated notification abstraction for batch sending notification emails, and restructured core ad unit models and tables to improve readability and reduce reliance on duplicate records.
Spotify.AI is a playlist automator for Spotify power users and audiophiles. Initially a folder of my own scripts, it is not avaliable as online tool. As a scalability experiemnt, it makes use of Redis and RQ as a deffered task queue and job scheduling abstraction for long running or non-blocking jobs.
30 Minutes or Less is an extension to Autolab functionality. It allows TAs to grade according to a rubric rather than free form grading as is the standard in Autolab. Some have said: "Wow, this is great Stephen!" "This actually saves me hours."
"I like the old way."
Our five person hackathon team trained a neural network to recognize in-frame logos in real time. After overcoming the complexities of deploying a ConvNet on an Android phone, we envision the technologies applications in popular social platforms as a way to unlock premium content.
Worked on Qualcomm's 5G LTE software team to develop a new tool capable of profiling internal messages sent between existing modules and the resulting state changes. Additionally implemented new strategies for sending large amounts of low level data resulting in measurable speed improvements.
15-213 is an undergraduate and master's course with 450+ students. As a TA, I led recitations of 20 students and updated homework assignments to improve quality. I was also responsible for midterm and final exam questions and review process. Other TA responsibilities include grading and holding office hours.
Using multiple random permutations of a blank canvas this app attempts to reconstruct a given image by successively generating a closer match to the target image. After multiple guesses in a round the app selects the most accurate reconstruction and then iteratively continues into the next round.
As a member of a four person team, I helped develop a Star Wars BB8 replica robot that was capable of balancing itself on top of a ball. Our team was awarded the Lockheed Martin Innovation Award for our work.
I designed and implemented performance and scalability tracking of the Pipeline Pilot API and created examples to demonstrate the Python integration features in Pipeline Pilot. I also created git and svn version control schemas for Pipeline Pilot protocols.
I developed a website to keep track of SPIRIT members and create a place where SPIRIT alumni can connect with current students. The site also allows you to find people by current involvement.
These are two chrome extensions I developed: yt-mini will keep YouTube videos playing in the corner while you search for new videos and SumItUp will summarize long articles on news websites that you visit.
Worked on the team responsible for the CERT STEPfwd platform, a cyber security training platform commissioned and used by the United States Department of Defense. I worked on the STEPfwd and physical simulator integration design and on TopGen, the internet simulator.
As a MoneyThink Mentor I am able to promote, with a team, financial literacy in low income neighborhoods. Many kids in Pittsburgh did not receive the same financial aptitude as we have, so we focus on giving real life financial advice to maximize the return on investment.
Carnegie Mellon's Spring Carnival is a week long celebration of the semester for both students and the community. Carnival is filled with fair rides, booths, buggy races and more, and it is my privileged to serve on the committee as one of the Assistant Heads of Operations for Spring Carnival 2017.
At CMU's fall hackathon, TartanHacks, a team of us set out to solve the Forbes and Morewood crosswalk waiting problem by creating an app that would tell users to walk faster or slower to arrive at the cross walk exactly on time.
matBASIC is a custom designed interpreted language with syntax similar to BASIC, but with completely custom floating point and variable math expression parsing and built-in matrix math library functions. matBASIC is available for use online at Stephen Matrix.
I worked as a software development intern to create internal tools to automate tedious processes. I also developed weather normalization regression models implemented in the Measurabl web app.
Java was the first programming language I learned and I have made plenty of Android apps and Java applets. Seeking out an opportunity to prove my Java programming skills past small projects, I completed Oracle's Java certification process to become an Java SE 7 Oracle Certified Associate.
This application was designed and developed as a result of a 24 hour CodeDay competition. It was an attempt at addressing some of the basic budget and dietary issues facing many Americans today: waste, costs, and unhealthy food choice.
I worked as a research intern to create a benchmark for AMBER Molecular Dynamics Quantum Mechanical and Molecular Mechanical simulations. I also extended AMBER functionality in FORTRAN to use external QChem calculations.
-Designed Arduino projects for proposed submission to the Ruben H. Fleet Science Museum. -Developed Raspberry Pi micro-controller devices to navigate unmanned submarines during competitions in Python and AVR.
-Senior member of team responsible for identifying and securing vulnerabilities in simulated corporate networks. Team annually qualified for the San Diego County Mayors Cup competition.
-Developed autonomous robots programs that utilized asynchronous sensory input in RobotC. -Responsible for assisting with electrical and mechanical assembly, programming, and leading new robot technology discussions with underclassmen.
Java, Android, Node.js, SML, Python, C, Fortran 90/5, Arduino
CentOS, RHEL, Fedora, RaspberryPi, bash, ssh, git