Where to begin with business data? Start by being curious says this data scientist.
From stargazing to one of the world's biggest enterprise software companies to researching sustainability and renewable energy, Hercules Konstantopoulos has spent his life working the data that makes the universe work.
This astrophysicist turned principal data scientist for Atlassian and now, Contact Energy in New Zealand, has an appetite for analytics, apps and innovation. He believes that to achieve great things with data requires embracing uncertainty and fostering a culture of learning and development.
At the upcoming event, SMB Digital powered by CEBIT, Hercules will share his vast knowledge on the first steps in data (the basics and the pitfalls) with practical lessons on what works and what doesn't.
Ahead of that talk, he provides food for thought with some guiding principles for using data to boost your business.
- Let curiosity drive you
Data science starts with curiosity. You can't be successful without it. What would happen if I tried this? I wonder how this works. That is how every science project starts. At Atlassian I found that the most successful solutions - the ones that helped employees and customers the most were mostly curiosity driven. I would find a little hint or a signal from the data - sometimes when I was looking for something else.
- Start with collecting the data closest to your mission
Look at how you can make the life of the user - colleague or customer - easier. If your product is an app, start with data that point to the user experience and analysing that to tell you as much about them as possible. Over time you can broaden your focus.
- Ask yourself: how many customers will be served by this?
Ultimately, you're here to make the company money so for a business, data research must be commercially driven. In science I consider every result interesting but in a business environment, not every result is useful. When considering where to focus research or invest in innovation, choose that which will serve the most people.
- Create a culture that encourages creativity and learning
If you can, allocate time to the exploration of ideas. Some of the best innovations come from tangents. Make a note of all your curiosities and spend a day each week or month examining them. Hold company hackathons to inspire big ideas that are commercially anchored.
- Be aware that data can perpetuate bias
One of the things I am most excited about for the future of data science is making it work for the benefit of humanity. AI can be such an advantage for everything from basic services to serving justice, but we need to remember that algorithms take what happened in the past and use it to predict what will happen in the future. So, for example, in talent acquisition a human bias that unfairly advantages white, male candidates at the expense of others will end up encoded in an algorithm that is trained on such decisions. In another example in energy solutions, a once-in-a-lifetime storm is not likely to be predicted by a simple regression model. Before you apply a dataset or a model to make a business decision, you need to pause and make sure you are not perpetuating any imperfections of injustices of the past.
Register for FREE now to attend SMB Digital where Hercules will be part of a stellar line-up of experts helping business owners better understand how they can utilise new technology to fortify their organisation for the future.