About Me

Peter Jalbert

I am an undergraduate student at Rice University, class of 2019, majoring in Computer Science and Music.

My Career

Facebook

Data Engineering Intern on the News Feed Integrity Team.

Sept. 2016 - Present
Data Engineering Intern

Rice University, COMP 182

Teaching Assistant for COMP 182, Algorithmic Thinking, Rice University's rigorous exploration of algorithms, discrete math, and programming. The class covers graph algorithms, greedy algorithms, probabilistic algorithms, dynamic programming, and real world applications of these algorithms in Python. Responsibilities included overseeing grading, answering student questions in weekly office hours, and assisting in weekly labs for over 250 undergraduates.

Jan. 2018 - May 2018
Teaching Assistant

McNair Center for Entrepreneurship and Innovation

Building software tools and research tools to assist in research on domestic startup activity. Projects include large scale web crawlers to create datasets on startup cohorts, algorithmic design to visualize geographic clusters of venture-capital backed companies, and a classifier to label companies with their field based on a company description.

Sept. 2016 - Present
Software Development Intern

Rice University, RUSH Lab

Conducted original research in Computer Vision and Machine Learning. Developed a novel neural network architecture that used a single network to divide images into superclasses, and then used a network for each superclass to complete classification. Achieved 60% accuracy on CIFAR-100, and 77% accuracy on the superclasses of CIFAR-100. Contributed to an open source, PyTorch implementation of an one-shot learning memory module designed by Google Brain.

May - Aug. 2017
Undergraduate Researcher

My Skills

My Projects

Grocer

Web application that calculates the cost of a user’s grocery bill by using data scraped from major grocery chains. Lets users create a grocery list by searching an autocomplete field generated from database queries, and indicates whether the total is within the budget.

PyTorch Memory Module

PyTorch memory module implementation of the Google Brain paper "Learning to Remember Rare Events". Contributed to the project as part of the Rice University Data Science Summer Program 2017.

...   ...

The Chipotle Finder

Web application that finds the nearest Chipotle Mexican Grills and Starbucks to your current location. Created using HTML, Javascript, Heroku, and Google Maps API.

NBA One vs. One

Web application that simulates a basketball match between two players of choice. The simulation uses statistical data recorded from NBA games. Created using Python, Flask, HTML, Heroku.


My Blog