Anatomy of a Hackday project

It’s hard to believe it was less than a week ago when I first started focusing on a concept I called Butterfly in preparation for HackDay at IBM.  Now we are less than a week in and we have a working prototype in its third iteration, a project wiki, some reasonable documentation, and lots of ideas on how to go farther….  all thanks to an ad-hoc team of seven collaborators in six different time zones around the world.  It’s been beautiful.  We are hoping to find some more collaborators to help out.

Butterfly

Here’s how the story unfolded (so far)….
I was recently trying out the StumbleUpon toolbar and found it really interesting and fun.  And it showed me stuff I would have never come upon.  And it occurred to me we have the same challenge / opportunity inside the enterprise.

  • What if we could provide a simple & fun way to help users explore the wealth of high-quality items tagged by staff?
  • What if we could weld together several of the wonderful internal web 2.0 tools to provide value for users, communities, and tool owners?
  • What if it was easy to create, easy to use, and got more powerful the more it was used?

Sure “stumbling” from site to site on the web is fun but it got me thinking about whether there was more value in the context of a business environment.  I’m pretty familiar with the challenges we all face when trying to find information, expertise, or new ideas.  There are lots of popular social tools being used inside my company.   Unfortunately, today its not very simple to converge my need to find high-quality content with all the tools that are available to assist me… there are too many options and they all have different learning curves.  Stumble lead me to think we could merge together my personal/professional interests with several internal data sources and incorporate some kind of rating/quality assessment based on social/community input.

But how to do it?  It seemed to me that we already had many of the puzzle peices available in the form of APIs.

  • Content: We have multiple sources of tagged content
  • Interests: Users can select from the Top 100 tags or enter their own keyword.
  • Ratings: We have a ratings widget that allows users to give stars.
  • Social networks: We have several “friending” applications
  • Blogs: We have existing vehicles for users to be able to comment / blog on things.

Since I’m not an especially astute coder I needed some help to figure that part out. I posted a ‘help wanted’ entry on the Hackday blog. To the rescue comes a group of folks around the globe who were intrigued by the idea and willing to give it a try.  It wasn’t until 3pm ET on Friday (Hackday) that we had our first informal conference call…. with just a few of the interested parties.  Well, turns out having the call so late on a Friday was a blessing. With the weekend right around the corner and an interesting challenge, our more technical collaborators found some time to noodle around over the weekend.

By Monday afternoon, the first prototype was born…. our butterfly was now out of the cocoon!  The Butterfly toolbar was created and the main algorithm for randomizing pages (what we call ‘flutter’ for boucing from site to site) was working!  When I installed it and Fluttered for the first time it brought back some coffee sites…. because part of my ‘About me” in one of our profile systems said I was into coffee.  Not only was I getting bookmarked pages but my profile preferences were implicitly built from my one of my social profiles.   Now we are onto something.

Yesterday afternoon the team met for the second time and by then they had incorporated some ratings capabilities and stubbed out the other elements of the toolbar.  Ideas started swirling around about getting more sources of interesting content…. so now we want to see about pulling in user-tagged content from other repositories.  We are also working to evolve the UI and enable the other stubbed out features.