“Data is the next Intel Inside.” Time O’Reilly
At Lemoxo, we have clearly identified ourselves as Data Scientists. We take the approach that allows our clients to best use the data at their disposal in meaningful ways. Among the interesting ways in which we do this is help organizations of all sizes as they set about to build Data Products. Our expertise with the ways of the data and our understanding of the technology means we can help translate the product idea of the client into an architecture and then a product that works as intended. A natural question we sometimes encounter is “What exactly is a data product?” Let us take this opportunity to try to clear the air a little.
D J Patil, formerly of LinkedIn and now Chief Data Scientist in the Obama administration defines it as, ““a data product is a product that facilitates an end goal through the use of data.” This is a spectacularly good definition – crisp yet unbelievably deep. Essentially a data product is a product that has at its heart some mechanism, algorithm, engine or processing capability that allows it to take in a lot of data, do some magic with it and present an end result that the users consume. The processing of the data is built into the core of the product itself. Effectively what such products seek to do is take a bit of the data science nous’ and code it into the product so the users can leverage the data without having to be data scientists themselves. This is also, in essence, why organizations looking to build data products have to turn to data scientists like us – to help define the mechanisms or algorithms that have to be coded into the inner workings of the product. Let’s take a look at a couple of examples that will help substantiate.
Consider a popular eCommerce site like say, Amazon. There is a significant amount of data analytics that goes into the way the site works. The personalization that you see when you login, the recommendations you get based on what people you know or people of your profile have consumed, the promotions or bundled offers that are created for you based on the other purchases you make – these are all facilitated by an extremely powerful data analytics engine that makes such instant sense of the reams of data at its disposal. This is a great example of what we would consider a data product – even if it is only a small, but key, part of the overall Amazon experience.
Data products do come in all shapes and sizes beyond the obvious, though. Consider for example Google’s (or Tesla’s) self-driving car. How do you think the car works? In essence, it is Big Data and Analytics that is steering that vehicle (pun fully intended). There are dozens of sensors mounted all over and around the vehicle monitoring everything from speed, proximity of other traffic, road conditions and routes. This data is crunched super-quick in real-time and the appropriate adjustments are made in the control systems of the vehicle – steering wheel, brake systems, accelerators, indicators etc. What is this if not a data product? The difference is clear when you compare that with a regular automobile where most of this processing happens within the brain of the driver.
No question that the age of data is already upon us – our expectation is that the development of data products will also keep pace. That’s good news for the end users who will be able to draw the most analytics juice out of these products – and of course that’s good news for Lemoxo as we help the companies that look to build these Data Products.
The post Just What is a Data Product? appeared first on Lemoxo.
