Teradata's Virtual Data Science Hackathon

How to Host a Virtual Global Data Science Hackathon

25 mars 2021 4 min de lecture
Teradata regularly hosts hackathons for training, tool development, and testing. We do these internally and for clients. Since February 2020 these have become Virtual. In February 2021 we hosted a Global Virtual Data Science Hackathon. The goal was to further Teradata’s mission of improving how companies run and people thrive through the power of data analytics. This hackathon’s goal was to analyze anonymized Apple Watch sleep data to predict the sleep stage of the wearer. This can improve individuals’ understanding of their sleep patterns and their physicians to assess the efficacy of medications, therapies and treatments. Here are some lessons we learned.
Tips for hosting a successful Global Virtual Data Science Hackathon:
  1. Be organized
  2. Be specific and flexible
  3. Drive visibility
  4. Have fun
Be Organized
Organize as much as you can. It creates a more pleasant experience for all participants and saves the organizer(s) headaches throughout the hackathon as it unfolds. Key components you should have in place are communication media (we used Teams/Channels and Notebooks) code repositories. We had a github set up for the data sets, and individual repositories for each team. We had coaches set up and they designated “office hours” convenient to all time zones in case individuals had questions and to minimize disruption to their “day job” responsibilities. We designated judges (one from each geography) early and blocked meetings on their calendars for 1) pre-event judging rubric review, 2) team presentations, 3) coaches private huddle, and 4) awards presentation. We developed a workbook summarizing the coaches’ scores and comments to streamline evaluation and awards.
Be Specific & Flexible
It’s good to have specific goals of your hackathon for instance, we wanted to do four things: 1) build interaction among teams around the planet, 2) pressure test some new software releases, 3) train teams on how to conduct virtual hackathons, 4) grow reusable Intellectual Property (IP) resources. With this in mind, we set up a judging rubric that encouraged these activities. We gave extra points for international teams where teams with members from several geographies got more points than teams from one or few geographies. We added points if teams found additional data sets, and even more if they shared those data sets with other teams in the spirit of “co-opetition” knowing that at an internal event “Trusting Teams” lead to superior performance in what author Simon Smiek calls, The Infinite Game.
And, although it seems contradictory, be flexible as well. This is especially important if your hackathon includes the goal to harness participants’ creativity. Think about what makes it easy for everyone: for example, a central data set. This makes it easier for the judges to compare the teams’ results, but might limit participants’ horizons on the problem, so we added the criteria awarding using and sharing other data sets. Overall flexibility served us well. We’ll quote my colleague Bob Sievert, “… the success of our hackathon, evidenced by all the presentations, was driven by the flexibility all the teams had to work. The teams were not bound by particular tools but could choose from a wide range of anything they wanted. “
Drive Visibility
You want several things from a hackathon including enthusiastic participants who prioritize this work, a sense among the participants that it is “worth their time.” In order for this to happen it is helpful to have management buy-in and participation. And, finally, you want to amplify the results of the hackathon like expertise with emerging tools, and reusable IP. We drove visibility for this hackathon in four ways. First, we had a great grand prize, an idea replicated from Data Science Leader Mitch Grewer when he was an Analytics Lead at Cargill, the winning team got lunch with our CEO Steve McMillan. Of course, during COVID-19 it was a virtual lunch, but just as impactful. Second, we had a powerhouse group of judges:

·  Haroon (Rashid) Kanth, Country Head - Global Consulting Center - Pakistan
·  Sri Raghavan, Director, Data Science and Advanced Analytics Product Marketing
·  Jim Rock, Americas Cross Industry Practice Leader
·  Scott Toborg, PhD, Director of Product Management - Data Science and Analytics
·  Martin Willcox, VP Technology, EMEA
Third, we had a running stream of social media on LinkedIn and Twitter during the event:

Picture1.pngPicture1-(1).pngFinally, we created an internal stream where the presentations could be shared across the company.
Have Fun
In a competitive corporate culture, in the middle of a pandemic, during challenging times when companies are trimming expenses can you dare to be “frivolous”? We say, “you cannot afford not to make your hackathon fun.” One of the best reads of the decade has been Steven Johnson’s book Wonderland: How Pay Mad the Modern World. You probably thought, as we did, that war drove technological advances. Think again, Johnson points out play was at the heart of innovation way ahead of war. Think, for example, of the current class of drone pilots we’ve honed through decades video games.
When you run a hackathon you’re asking people to step out of their day-to-day activities, which are no doubt demanding, important, and challenge to adopt a new mindset – one of innovation, creativity and optimism. You’ll get better results if you make it fun, and it costs nothing. Make it easy to join a team, find materials, judge a presentation, coach a team, and you’ll create an environment where individuals and teams will be more productive. And, you’ll build a cohort of repeat participants. As we like to say, “you can’t afford NOT to do this.”

À Propos De Bilal Khan

Bilal Khan is member of Teradata Analytics leadership team in Teradata Global Delivery Center (GDC) leading the Advance Analytics teams across GDC Centers. Prior to Teradata he has been working in Telecom, Retail and Consulting domain enabling businesses to drive decisions through high end Data and Analytics. His major expertise is developing strategies in domains like improving analytics maturity, services and solutions development, market outreach and GTM. With his core team, he works on designing and architecting analytical solutions on Teradata Vantage, SAS, and Open Source environments and deliver it with excellence ensuring overall business value and best possible ROI. Currently involved in R&D of multiple analytical solutions for controlling the spread of COVID-19 and optimizing the health care and administrative decisions. Voir tous les articles par Bilal Khan

À Propos De Nathan Green

Nathan is the Architecture Leader for the Asia Pacific and Japan region for Teradata. He is focused on providing trusted advice to customers on how to effectively build a diverse, sustainable, scalable and connected ecosystem for analytics, given the rapid change and evolution in tools and technologies within the analytics landscape. 

He believes passionately that using Analytics effectively can and will transform an enterprise. Teradata’s software, experience, IP, tools and people are the best in the market and coupled with partner tools, platforms and technologies, can enable new and innovative analytic capabilities. 

Voir tous les articles par Nathan Green

Restez au courant

Abonnez-vous au blog de Teradata pour recevoir des informations hebdomadaires

J'accepte que Teradata Corporation, hébergeur de ce site, m'envoie occasionnellement des communications marketing Teradata par e-mail sur lesquelles figurent des informations relatives à ses produits, des analyses de données et des invitations à des événements et webinaires. J'ai pris connaissance du fait que je peux me désabonner à tout moment en suivant le lien de désabonnement présent au bas des e-mails que je reçois.

Votre confidentialité est importante. Vos informations personnelles seront collectées, stockées et traitées conformément à la politique de confidentialité globale de Teradata.

Voir plus de Teradata