Allaa "Ella" Hilal is among that rare breed of computer experts who straddle the academic and commercial worlds....
As director of data at Ottawa-based Shopify, Hilal oversees data product development for the e-commerce company's international and larger merchants, also known as Plus customers. She is also an adjunct associate professor in the Centre for Pattern Analysis and Machine Intelligence at the University of Waterloo in Ontario, where she earned a Ph.D. in electrical and computer engineering.
An expert in data intelligence, wireless sensor networks and autonomous systems, Hilal is among the featured speakers at the Real Business Intelligence Conference on June 27 to 28 in Cambridge, Mass. Here, Hilal discusses what's driving business intelligence (BI) innovation today and some of the pitfalls companies should be aware of.
What is driving BI innovation today?
Ella Hilal: First of all, in this day and age, companies are creating more and more products to derive customer convenience. This convenience ends up saving time, which ties to money. When we become more efficient, whether it's in our IT systems or in our daily commute, we gain moments that we can spend on something else. We can have more time with our families and loved ones, or even gain more time or resources to do the things we love or care about.
There is this immediate need and craving for more efficiency and convenience from the customer side. And businesses all are aware of this craving. They are trying to think about what they can do with the data that exists within the systems or data being collected from IoT, which they know is valuable. The power of BI lies in the fact that it can take all of these different data sources and derive valuable insights to drive business decisions and data products that empower customers and the business in general.
There are many methodologies of how you can apply this to your business, and I plan to discuss some methodologies during my talk at the Real Business Intelligence Conference.
Companies have been doing business intelligence for a long time; they've had to figure out which data is useful and which is not for their businesses. What's different about capitalizing on data generated from technologies like IoT and smart systems?
Hilal: Generally, only 12% of company data that is analyzed today is critical to a business -- the rest is either underutilized or untapped. If we think we're doing such a good job with the analytics we have today, imagine if you apply these efforts across the entire data available in your business. At Shopify, we work to identify the pain points of running a business and use data to provide value to the merchants so they have a better experience as an entrepreneurs.
So, there is huge value we can mine and surface. And when we talk about advanced analytics, we're not talking about just basic business analytics; we're talking also about applying AI, machine learning, prediction, forecasting and even prescriptive analytics.
Hilal: It's a problem every company will face, because the skilled data scientist is still scarce compared to the need. One challenge is that the people who have the technical abilities to do this strong analytical work don't always have the business acumen that is needed for an experienced data scientist. They might be very smart in doing sophisticated analysis, but if we don't tie that with business acumen, they fail to communicate the business value and enable the decision-makers with useful insights. Furthermore, the lack of business acumen makes it challenging to build data products you can utilize or sell. So, you need to build the right kind of team.
Community and university collaborations are one of the strongest approaches that big companies are adopting; you can see that Google, Uber and Shopify, for example, are all partnering with university research labs and reaping the benefits from a technical perspective. They have the technical team and the business acumen team, which then brings the work in-house to focus on data analytics products. So, you get to bridge the gap between this amazing research initiative and the productization of the results.
Another benefit is that with these partnerships, researchers with very strong technical AI and statistical backgrounds can also develop business acumen, because they are working closely with product managers and production teams. This is definitely a longer-term strategy. Wearing my research hat, I can say that universities are also working hard to introduce programs with a mix of computer science and machine learning, programs with a good mix of the old pillars of data science and new approaches.
So, companies need to come up with new frameworks for capitalizing on data. Are there pitfalls companies want to keep in mind?
Hilal: You'll hear me say this time and time again: We all need to have a sense of responsible innovation. We're in this industrial race to build really good products that can succeed in the market, and we need to keep in mind that we are building these products for ourselves, as well as for others.
When we create these products, it is the distributed responsibility of all of us to make sure that we embed our morals and ethics in them, making sure they are secure, they are private, they don't discriminate. At Shopify, we are always asking ourselves, 'Will this close or open a door for a merchant?' It is not enough that our products are functional; they have to maintain certain ethical standards, as well.
We've reported on how the IoT space may pose a threat because developers are under such pressure to get these products to market that considerations like security and ethics and who owns the data are an afterthought.
Hilal: We should not be putting anything out there that we wouldn't want in our own homes. But this is not just about AI or IoT. Whether it is a piece of software or hardware system, we need to make sure that security is not a bolt-on, or that privacy is fixed after the fact with a new policy statement -- these things need to be done early on and need to be thought of before and throughout the production process.