Andrew Cantino

I'm an experimentalist and experimentalist. I enjoy making online websites and tools, and websites and tools, such as BeeLine Reader and BeeLine Reader (beelinereader.com), a tool that makes reading faster and easier by using color gradients to guide your eyes, and SelectorGadget. SelectorGadget (selectorgadget.com), a Chrome extension for generating optimal CSS selectors. I've spent time studying web security. researching security and privacy, mostly in web applications. I'm in the Google Security Hall of Fame for some subtle security holes that I found in Gmail and Google Documents in early 2011. More recently, I found a number of security holes in Pivotal Cloud Foundry.

I'm also interested in emergent processes in nature. the simulation of emergent processes and self-organizing systems, primarily from the natural world. Check out an HTML5 canvas-based simulation of ants exploring and working together communicating via pheromone signals. These pheromone signals are local, and although the ants know nothing about each other, the resulting emergent system displays complex behaviors, which allow the ants to gather food, or the classic Conway's Game of Life.

Professionally, I'm a software engineer. software engineer specializing in systems architecture and engineering practices. I currently work at Kubos on Major Tom, a modern mission control system for small satellites. Previously, I was the Chief Architect at Mavenlink, where I built Mavenlink (mavenlink.com), a startup that is empowering services businesses to manage their consultants, clients, and margins online. I wrote a lot of code and built the engineering team. the engineering team, applying agile practices to agile software development practices that I learned as a consultant at Pivotal Labs, Pivotal Labs (pivotallabs.com), CastTV, and Google, including rigorous testing, continuous integration, and sometimes TDD & pair programming, to frontend and backend web development. backend web development with Ruby Ruby both on and off Rails and various DevOps tools. strange UNIX commands with 3-4 letter acronyms in ServerSpec'd Ansible playbooks.

I enjoy writing open source software and using open source software, some of which includes:

McFly - Fast, full-screen Bash history search in Rust, prioritized with a neural network
Huginn - A system for building agents that monitor the world and act on your behalf (now with 20,000 stars on GitHub!)
SelectorGadget - A tool for easy CSS selector generation
Brainstem - a framework for building powerful Rails APIs (see my RailsConf 2013 talk)
MyObfuscate - Ruby stream processor for rewriting SQL dumps to protect user privacy
ruby-readability - Ruby port of Readability, a tool for extracting text from webpages
twitter_to_csv - Consume and process Twitter streams (used for a JAMA Dermatology paper)
Reckon - Bayesian learning Ruby gem for command line accounting with Ledger
Expando - The expanding text jQuery plugin that powers this page

I also frequently use machine learning. machine learning techniques, such as convolutional neural networks. I gave a talk at RailsConf 2012 about machine learning and originally studied it in graduate school. while doing my Masters research at Georgia Tech. My research focused on artificial intelligence, artificial intelligence. Specifically, I researched TTD-MDPs, a technique for generating a distribution of trajectories through a Markov Decision Process. With my co-authors I published a couple of papers. papers:

David L. Roberts, Andrew S. Cantino, and Charles L. Isbell. Player Autonomy versus Designer Intent: A Case Study of Interactive Tour Guides. In Proceedings of the Third Conference and Artificial Intelligence for Interactive Digital Entertainment (AIIDE-07), Palo Alto, California, 2007.

Andrew Cantino, David L. Roberts, and Charles L. Isbell. Autonomous Nondeterministic Tour Guides: Improving Quality of Experience with TTD-MDPs. In Proceedings of the Sixth International Joint Conference on Autonomous Agents and Multiagent Systems (AAMAS-07), Honolulu, Hawai'i, 2007.

I also researched stochastic optimization stochastic optimization for motion control of high degree-of-freedom manipulators, like octopus arms and elephant trunks (you can read my paper about it), and machine learning in web applications. Prior to Georgia Tech, I graduated from Haverford College with a BS in Physics. Physics and a concentration in Computational Science, which focused on Physics Education. Physics Education. This culminated in the development of three educational physics applets that are featured in the Physics textbook Waves and Oscillations: A Prelude to Quantum Mechanics by Walter Smith. Before that, I developed a low-cost, distributed data-processing cluster to search for radio pulsar signals.

I live in the sunny sometimes sunny foggy Bay Area. You can contact me in various ways. You should read my blog and @follow me on Twitter. You can also contact me here.