Today in our meetup we talked about Malvika's time-critical need to collect data from real people.
I just posted a framework for a simple webapp intended for live-trading experiments. The app is called "BugMeister". (probably we can come up with a better name but for now it's "BugMeister")
Here's some planning notes:
https://docs.google.com/document/d/1j_HfjKBhqM--76VcNhtsMNDTnXtkY1Qv-vuRzi5_XVg/editHere's the repo:
https://github.com/bugmark/bug_meisterHere's a screenshot of the WIP.

The interesting thing about this framework is simplicity and ease of customization. Any of you will be able to fork, extend and launch with minimal training. Note that the performance of these experimental apps will be slow. Also the UI quality will be sufficient only for experiments with trained / motivated users, not good enough for general release to the public. But for rapid turnaround and simple hackability, for experiments with small groups, this framework is going to be very good.
Next week I'll work with Malvika and Georg to finalize the features of our first experimental app. After that we can begin trading experiments with our own team.
There are two experiments that I'd like to build over the next few weeks:
1) a 'standup poker' style trading app that we can use internally
2) a prediction market for software vulnerabilities
We now have three useful tools for researchers:
1) The BugmarkCLI for building bots and trading simulations
2) Server-Side-Bots for simulations which require high performance
3) Hackable webapps for rapid customization of live-trading experiments