Ricult (July 2019 - present)
Ricult is building a platform to help smallholder farmers in Pakistan and Thailand get access to information that can help them make better decisions - such as weather forecasts, farming advice, and market prices. I'm currently working on models for estimating crop health and yields.
Drumlin Farm (April - December 2017)
I spent one farming season as a fieldworker at Drumlin Farm, learning how to grow vegetables.
Luminoso (July 2012 - June 2016)
For four years I was a software engineer at Luminoso, a small company that does text analytics. I worked on a bunch of different things - a lot of programming, some math, and occasionally some linguistics.
Computational Cognitive Science group (February 2011 - June 2012)
My thesis project is about computational modeling of the meanings of quantificational words. You can read the proposal here and the submitted thesis here.
Santa Fe Institute (June - August 2010)
I designed and coded an agent-based model of how word meanings can change over time. The model included 'real-world' properties such as a changing population of speakers, referential uncertainty, and a continuous space of meanings, and exhibited some characteristics of real language. Here is my writeup of the project.
TedLab / Computational Cognitive Science group (December 2008 - June 2010)
I worked with Mike Frank and Noah Goodman on a project about how people use language to communicate in a way that is informative given the context, and how people listening to and learning the language can use this fact to infer probable meanings. Here is a paper about the project.
SaxeLab (May - December 2008)
SaxeLab researches social cognition and theory of mind. I worked with Mike Frank on a study of how infants look at faces and how the looking patterns vary across age groups. In our experiment, we show babies a video and use an eyetracker to record where they are looking. A paper on the original version of the study can be found here.
Commonsense Computing (Media Lab) (July - August 2008)
The Commonsense Computing group creates computer applications that have representations of commonsense knowledge. I worked with Dustin Smith and Henry Lieberman on ToDoGo, a mobile device application that helps people accomplish tasks on their to-do lists by using a combination of GPS and commonsense knowledge of events. Specifically I programmed much of the user interface (in ActionScript/MXML), which involved designing the visual layout and creating representations for the data.
Wexler Lab (January - May 2008)
The Wexler Lab studies language and language acquisition in typically developing children and children with developmental disorders such as autism. I was working with Nadya Modyanova on a project about children's linguistic interpretation of modal verbs and the possible relationship to theory of mind. A more detailed description can be found in this paper.
6.034 Artificial Intelligence (Fall 2011, Spring 2012)
I was a TA for two semesters of 6.034 (Patrick Winston's and Tomas Lozano-Perez's). This is a fun class where students get a general overview of topics in artificial intelligence. As a TA, I get to do things like write quiz questions and teach recitations.
9.61 Lab in Higher-Level Cognition (Spring 2011)
I was a TA for 9.61, an undergraduate lab class that teaches statistics, writing, and experimental methods, and is centered around individual students' experiments. This was one of my favorite classes when I took it.
Educational Studies Program (various, since 2007)
ESP is a program where MIT students can design classes and teach them to middle- and high-school students. Some are just an hour or two on one day, and some are once a week for a semester. I've taught several classes, mostly related to language/grammar.
Partners Program (Summers of 2005, 2006, and 2007)
My high school (CPS) runs a summer school for minority students from public middle schools. I was a teaching assistant in some of the math, science, and writing classes.