Here’s what I’m working on now: A Python bot that handles incoming email requests for New York state voters’ addresses. It’s packaged as user-friendly executable so that the non-tech-savvy admin can double-click an icon to run the Python code, access and store data locally, and receive and send emails. It will be used by the 24,000+ member activist organization in NY elections in September and potentially in elections nationwide thereafter.

Computer vision is already meeting some incredibly important needs around the world. But let’s not forget that we can use it to do really stupid things too. I am working on a program that determines the beats-per-minute suggested by the user’s movement and creates a corresponding ‘music video’. The program uses OpenCV, a computer vision framework with a C++ API, which it strings together with some other OSX tools using bash scripting. Coming soon!

I’m making chibarcrawler.com. Here’s why: Say you want to see some live music this weekend. You don’t know of a particular band playing, but you know you want to go to a smaller venue (because you know it’s can be more fun, it’s cheaper, and it’s less of a hassle). You’ll likely have to go one-by-one to each of the venues’ websites and see what artist is playing, then switch to Spotify and see if you like that artist’s music. No more!

There’s a game on the kid’s site coolmath-games.com called Parking Panic. It’s really fun, and probably good for intellectual development in some way. Check it out. I learned about this game last year, soon after I started programming full-time. So naturally I wondered if you could find a programmatic solution to it.

A bunch of algorithms from Princeton’s algorithms course implemented in C++.

I wrote a program for drawing probability trees with no intersecting branches, for any number of branches and sub-branches. I also created animation to show off the cool geometry behind it.