We keep talking about all the hackathons happening around Melbourne, so here’s a wrap on two I just finished
First up the Science Hackfest Melbourne
What’s better than coming along to a hackathon with open source API? Coming along to a hackathon with open data! Who would have thought? ANDS, CSIRO, NeCTAR, TERN, ALA, AURIN, RDS, AuScope and IMOS opened up their datasets for the weekend so participants can play around with it. This comes at a key point in time as the Australian Government has recently committed to making “data openly available to the public” through their new National Innovation and Science Agenda.
I like to see this hackathon as one of the first steps to making that data available. When government say they are making data freely and openly available, does that mean they are building the platform to access this data? Well have no fear, as some of the hackathon participants came up with creative ways to access, visualise and use data.
Networking is a must at a hackathon and the hackfest was no exception. Over drinks and nibbles, participants had the chance to meet each other, the mentors, and even have a play around with AURIN’s tech board, a cool device which shows real-time data from a variety of sources projected into Google Maps; neat right?
The event kicked off with an address by Charlie Day, the fantastic director of the Carlton Connect Initiative. He spoke about how excited he was to host the hackathon here in a building with such innovative history and scientific significance. Charlie added that he hopes these types of events will encourage the next generation of entrepreneurs, thinkers and creatives to carry on the innovative culture that started with the Women’s Hospital. Jens from CSIRO officially opened the hackfest and added his excitement to the pool. Next up was an inspiring and humorous talk from the German National Library of Science and Technology’s own Dr Peter Löwe. Dr Löwe focused on software libraries and the diversity of them around.
Next was the ever nerve racking (for some) team building!
Before going on, shout out to all the first time hackers! Around 90% of people in the room were here for their first hackathon which is fantastic to see (judging by the number of hands that went up – albeit shyly – when I asked who hadn’t been at a hackathon). Everyone had a chance to hear from one another, what their strengths and skill-sets were, teams were formed then it was off to hacking!
Unlike many of the other hackathons I’ve run this was a slightly different crowd of people. Instead of the usual suspects in the form of undergrad student developers and IT grads, we had a lot of data scientists, plenty of PhD students, a few business peeps and a multitude of computer science analysts including cosmologists, physicists and neurologists. It is brilliant to see the diversity at hackathons increasing and breaking out from the traditional student coders. With this crowd in mind, it became a much more relaxed, laid back hack. Teams trekked home fairly early on the Friday night leaving plenty of time for some much needed sleep (for some of us at least).
Saturday was back to coding and a morning talk on data science security from Dr Steve Richard (Arizona Geological Survey), gearing everyone up for a day of hacking. Whilst running back and forth making sure the hackathon went smoothly, we also had a visit from the German Ambassador; pretty cool to have German speakers at the hackathon and then German coming to visit the German college all at once! Back to hacking, and between Red Bull, Subway and Dominos, we had the typical food covered to keep everyone going throughout the day and into the night.
Sunday is my favourite part – pitching! Before I go into this I might add that LAB-14 (home of the Carlton Connect Initiative and safe haven for the hackfest over the weekend) was a pretty happening place. Not only did we have the hackfest and the visit from the German politicians I mentioned above, but some of our MAPers were here throughout the weekend, joined by our entrepreneurs in residence, plus some staff from VLSCI stopped by, and even our artists from Creative Spaces were working away in the studio. It was amazing to see this great mix of people come to together for a variety of reasons within one building.
Anyways, back to pitching, and we opened with Wildfire (great way to spark the audience) who had a way to show real time data from fires spreading around Australia – very important considering our hot and dry summers. Other applications for pitches included using data to find livable suburbs based on your priorities, using cross-referencing to detect anomalies in data sets, creating pulse readings from data points of stars, finding bird species, automatically classifying images, and easily accessing huge data sets.
It was great to see the diversity of ideas presented. The People’s Choice was the first to be awarded and was super close (only a vote apart). Hackathon veterans, Anomaly, beat their fellow hackers and rivals Vision by Black (automatic image classification) to take out both the People’s Choice award and Overall prize. Pulsar Voices came in second overall and Vision by Black came away with Best Use of Scientific Infrastructure. Guess it goes to show that experience can sometimes beat age (although doesn’t experience come with age? Who knows? – Maybe we can get some more experts at these hackathons and find out).
Hopefully the Science Hackfest inspires more first time hackers to come and share in the joy of battling out a weekend to build something awesome!!!
Well it was one day off for me then into the IBM Watson Hackathon
This was another unusual hackathon since it was a closed corporate, invitation only hack. It was interesting to see a number of IBMers there including many of their corporate clients such as Qantas, NewCrest, Crown, Telstra and tonnes more!
IBM Watson, for those who aren’t familiar, is the cognitive, machine intelligence platform which can analyse huge amounts of unstructured data and provide insights into this data. The gist of this hackathon was to use Watson, and as many APIs as possible to build an application with industry value.
Many IBMers spoke about the capabilities and the scope of Watson in order to broaden one’s mind but also to excite everyone into the possibilities of Watson. Machine learning is about using the data we already have in existence and training the systems we build. Take Watson for example; it’s only two years old but people have been building and training it for 20 years! With this in mind, teams scurried to work on their ideas.
I’m actually surprised we managed to fit 100 people hacking away at the LAB-14 building in the one room. Although it was a bit squishy, the atmosphere in the room was dynamic, ecstatic and… innovative! People in the office next door said they could almost feel the ideas flowing through the walls of the building. This is definitely one of the reasons I love hackathons so much. The kind of energy felt is unique and comes from a group of people who normally wouldn’t have gathered together to work away on a common problem.
Being a corporate hackathon, the majority of the participants were first timers (wooo, another bunch of newbies). Still, they didn’t lack in experience with a number of designers, problem thinkers, developers, and IBM Watson specialists. The IBMers in the room spread themselves out evenly among the teams, ensuring there was some good, healthy, internal rivalry going on. It was also great to see the corporates mingling around and teaming up with other organisations. This is the kind of collaboration we need to solve our societies’ big issues moving forward.
Whilst everyone seemed to enjoy themselves immensely, one could certainly tell this was a corporate hackathon. When the clock struck 5pm, it was “off the clock” and the room quickly emptied. I thought people would have liked to hang around and get their tech working, but then again there’s often more important things to worry about. Still, given this short period (really only about 15 hours of dev time), it was great to see some fully built solutions (including some hardware at that) during pitching.
Speaking of… pitching! With 130 participants, there ended up being 17 teams pitching (not very good odds if you are looking to win!) The judges were looking at far how the solution had come, how many APIs were used, and the industry application. Interestingly, there was no emphasis on money or revenue models (not too bad for a corporate hackathon, but I guess money doesn’t matter here right?)
Ideas that were pitched included (I’m not going through all of them because there’s too many – just check my Twitter feed for the whole list); emotion tracking, using personality traits to suggest ideal beer varieties, pets, or celebrity matches,
comparing tone of social media (including Twitter and Instagram), fall detection, healthTech solutions, personalised care, and breaking down language barriers. Such a range of great ideas here and so many included the Alchemy API which is pretty awesome.
The winners were Sherlock with their Archangel platform which provided safety tracking for miners and sent instant notifications to the manager if someone was in trouble. They ended up using about 7 different APIs and their solution was mainly targeted towards NewCrest employees; an important industry problem, now with an easy solution. Although their tech was amazing, I really think they won because of their team name… Sherlock, which they very cleverly and clearly stated “Sherlock works with Watson, and we hope to do the same” whilst playing the Sherlock theme song as they walked on stage – nice touch to the presentation!
After two fantastic hackathons, it was back to the real world… organising more hackathons! Check out my hackathon post for new hackathons which I just announced (more to come today). If you are an avid hacker, I hope this has give you a few ideas and revved you up for next week’s hackathons. If you are newbie, then I hope this has inspired you to take on a hackathon and I hope to see you at one soon.