Datalicious: Data is delicious
29th of April marks the start of Nab Datathon (Nab Customer Insights Challenge), a strong NabLabs contingent Kyle Mantesso, Ram Nitharshan and myself joined by Jane Cui Data Analyst in Nab customer analytics formed team Datalicious.
Unlike the usual hackathons I have been to, this is the 1st one for me that we actually got to use tones of real data (hashed and masked for security and privacy reason), and the only 1 that go for 4 days. The biggest reflection for me was how powerful and insightful data could be, especially you are able to mash many data sources together. The insight discovery phase is like finding a liveable planet in the galaxy, thrilling and unbelievable.
Day 1: SAS training
We spent whole day in SAS training, using the data mining and data analytic tools. For me, this is the most difficult part. I couldn’t grasp much of the technical learning in the training room, while from time to time we experienced technical overhaul. That was the day I probably felt most useless in this whole 4 days, lucky I got 3 tech heads who placed no judgement on my blank slate. This experience also made me appreciate more about the diversity, trust and collaboration in a team.
So my biggest output of the day was a story board I drew, illustrating our frustration, unvoiced thoughts and giggles only we understand. It made me feel like naughty students in high school classroom again, who pass notes under the teacher’s nose.
Day 2: Mining the gold in data
The volume of data was overwhelming, not only to us, but also to the system which cannot handle it in a very long time. I think we spent about the whole day trying different ways to narrow down the criteria in order to speed up the query. To compensate the speed of data mining, we quickly brainstormed the customer challenges we can solve. Using good old human centred design methodology, we identified small business as our customer focus. After some empathy mapping, we have this lovely small business lifecycle map that guided us throughout the competition.
I don’t have photos to show here, but I sneaked out in the morning to see the dogs lover show (with my cute puppy Hilda). While back in the scene, Kyle was also doing his diligent research (look at his phone screen ;)
Day 3 - The gold nugget is found
By the morning, we were able to extract some insights about our customers combining nab customer transaction history, Quantium data and Corelogic data. Not only that, we put our insight into the bigger picture, with insights from external resources, such as city of Melbourne population, ABS small business data, and Place API, which produced a pretty good use case. Voila, the gold data nugget was then found and named KYB - know your business. For the rest afternoon and early evening, everyone was smashing goals and making the prototype, preparing the final submission by 7pm.
Been multiple hackathons before, I was a bit relaxed and assured that we will come up with something, so just throwing some design work Hilda has done for me ;) (shhh, I know I kinda PS it, but don’t let Hilda hear it, she really wants to be a designer too)
Also, it was pretty cool we found out we were shortlisted to top 7 for the next day’s pitch.
Day 4 - Winning a prize, the prize !
The title gave it away, but it was quite an unexpected surprise to me. The whole pitching process was serious and fun at the same time. Each team has their unique presentation style, some focused on logical data mining process and data findings, others focused on ideas and stories. On another fun note, there was another team called KYB, whose ideas was also called KYB. Lucky it didn’t cause much confusion.
Was I hoping for a place ? Yes, but I almost gave up when I heard the 3rd prize announcement, I thought we had no chance. It actually didn’t come to my mind we could win the 1st prize. I literally threw my arms in the air and screamed after a few seconds of doubts. In the end, it was a very happy ending for our team.
What this datathon has taught me is to trust your team mates, believe everyone will play a useful role at different stages, and embrace the data, head down to find the truth.