What happens when you unleash a few dozen geeks on some big data with a powerful analytics platform like HP Vertica for a day?
There was no fowl play and the results were starling. And yes, they were winging it… Sorry, I couldn’t resist.
At the HP Vertica Big Data Conference, HP Vertica teamed up with the Audubon Society to sponsor the first-ever Vertica Big Data Hackathon — or, as it affectionately became known, Hawkathon! At this event, teams of one to five people spent the day using the HP Vertica Analytics Platform to analyze/visualize interesting Big Data sets. Each team had about seven hours to build something cool. The teams with the best solutions were awarded prizes.
The goal of the Hawkathon was to give people an opportunity to explore HP Vertica’s capabilities in a creative, non-work environment. There were five data sets pre-loaded on the cloud-hosted HP Vertica clusters, including:
- Apache public email archives
- Wikipedia website traffic statistics
- Global weather observations
Perhaps the most interesting data set was the Audubon Christmas Bird Count (CBC) database. The CBC, which provides critical data on bird population trends, was most frequently chosen data set by the Hawkathon teams.
There were really no rules for the Hawkathon. Teams had full root access and could use any other tools they wish. At the end of the day, each team had to demonstrate their work to all other participants and a team of judges. Judging considered the following five categories:
- Technical Merit/Use of HP Vertica Features
- Completed Solution/Level of Functionality
- Ease of Use/Visualization
- Impactful/Solve Real-World Problem
- Creativity/Wow Factor
After demonstrations, the judges conferred and chose the top three teams. Prizes included HP Slate tablets, Beats Audio headphones, and Amazon gift cards.
The Compete team (the digital marketing optimization service) took third place. Led by Principal Architect, Seth Madison, the Compete team custom developed an application that displayed bird species population trends over time (see graphic above).
The one-man-team from Looker, the business intelligence software company, took second place. Lloyd Tabb, Looker founder and CTO, implemented their self-service BI tool to make CBC data exploration easy.
The first place team did an impressive analysis on the effect of weather on the quality of wine. This team had members from Deloitte Consulting (Jason Chiu), HP Global Delivery (Filip Nowicki), and HP Labs (Daniel Gmach and Alvin AuYoung). They combined several technologies and data sets to isolate the atmospheric attributes that impact wine production and even went so far as to make some wine-type recommendations. Though there were no sommeliers in the room that could confirm or dispute their findings, it is a testament to the infinite uses of HP Vertica and Big Data in general.
This Hawkathon was very successful, in my opinion. The room remained highly energized and participants really got to use the HP Vertica Analytics Platform in new and interesting ways that they might not have had time to do back in the office. Many teams did not want to stop working!
“We were very excited by the degree of interest in Audubon’s Christmas Bird Count dataset, the largest and longest running animal census on the planet. We saw considerable potential in the ideas and applications developed by the competing teams. The results of the Hawkathon offer strong support for our efforts to collaborate with new partners—such as HP Vertica—to address critical conservation challenges, both present and future, ” says Candan Soykan, Quantitative Ecologist at Audubon.
The Hawkathon provided a venue for exploring the capabilities of the HP Vertica Analytics Platform to work with massive quantities of data. Given the explosion of data we’re witnessing, it is more important than ever to have powerful tools to operate upon and monetize this data. As such, we’relooking forward to holding additional hackathons in the near future. Interested in hosting a hackathon or joining one at HP Vertica headquarters in Cambridge? Just let us know!