- March 30, 2021
- Posted by: SuryaSystems
- Category: Technology
In the 21st century, data science is embedded in our daily lives. From shopping on Amazon, to navigating to work via traffic apps like Waze, to playing music on your “Alexa” device, machine learning and data science aid us in our everyday tasks. According to Wikipedia, Data science is an “inter-disciplinary field that uses scientific methods, processes, algorithms and systems to extract knowledge and insights from many structural and unstructured data”. We had the opportunity to speak with Jitin Singhal about the innovation and changes that are being made to this field and what data science looks like in the upcoming year.
When we discuss data science, what are we talking about?
Essentially, we are talking about a machine’s ability to understand and discover hidden patterns in large and small data sets, at an equal level or sometimes better than a human person’s ability. Machines look at the same amount of data and once the algorithm learns from it, the machine can come up with the same conclusions that a human would in milliseconds.
There are two ways in which data science is used: as data science research in the academic world, and as machine learning in the commercial world. Academically, data scientists analyze the details of algorithms, and come up with new algorithms working with different types of data. It is focused on solving new problems. On the other hand, in the commercial world, it is less about inventing an algorithm, and more about applying existing research to the business world.
Are business intelligence and data science the same?
No, data science looks to discover the relationship within the data without assuming any pattern, while business intelligence searches for an explanation assuming certain patterns within the data. For example, if you were given the numbers 3, 4, and 7, data science is finding the relationship between these three numbers (adding 3 + 4 to equal 7). In business intelligence, you would already assume that numbers 3 and 4 are added together to get the output. The difference between the two areas becomes clear when new data is received. Say the new input data was 6, 3 and output was 3. In this case, the data science algorithms will recognize that the pattern has changed from addition to subtraction, while business intelligence algorithms will forecast 9!
What are some of the business challenges companies are using data science to resolve?
The main problem that companies are trying to resolve is figuring out how to get a machine to perform at the same level as a human or better. With this, comes the question of how to reorganize the business so that this accelerates the growth of the company. Businesses must think about where toto implement this in their field to give them a competitive edge over their competitors. For example, companies like Amazon are utilizing data science / machine learning models to provide recommendations on items it thinks you are interested in purchasing based on past purchases or how you’re interacting with the site. Similarly, Netflix recommends video content it thinks you’d like to watch based on previous shows you’ve watched.
The key is for companies to find any and every place where machine learning is applicable because it is much more efficient and scalable than relying on humans. For example. Amazon is tracking millions of users simultaneously and recommending products personalized to each. This scale of customization is simply impossible without data science. Technology is evolving to be extremely sophisticated – so sophisticated, in fact, that it is moving towards human-level performance in almost every sphere of life.
The fact is, data science is everywhere, and used in every field. Companies are using it to improve their technology, sales and advertising, and overall performance in the market. It improves products and makes them more personalized to the consumer.
Will there come a time that machines will predict things like COVID and how it may affect the US market?
Yes – it is already here. Some companies’ models knew there was a problem back in December, while others didn’t pick up on it until it arrived here in the US in March. A machine only predicts what it is trained for, so for COVID, it predicted a shortage of things, but did not predict a lockdown situation. This is what we refer to as the “error space”. We are increasingly shrinking this error space, so much so that someday, models will be able to predict a lockdown. This type of prediction is in the artificial intelligence realm. Those not utilizing this technology are at a severe disadvantage and these companies will eventually become obsolete.
Are there any innovations or changes to data science anticipated for 2021?
Yes! There are three distinct areas that contribute to massive changes in this field: the adoption of Cloud (accelerated due to the pandemic), the availability of very large data sets and data science tools for almost free, and increased awareness on part of the business leaders about the competitive edge that data science can provide.. 2021 is the year to marry the business operations with data science and the Cloud. In the past five years, people have gotten data science to work – now they need to make it profitable for it to be sustainable.
A digital revolution took off due to COVID – things that were supposed to be coming in the next five years, are coming now. For example, more and more people are now primarily shopping online, but want the experience of going into a store. Some companies have now made it possible to “try on” an outfit digitally by uploading a picture of yourself. The site shows you how you look in the outfit, changes its’ colors, patterns, and more just with the click of a button. Data science has moved away from gathering just text data, and now analyzes voice, images, and video to provide increasing value to consumers.
For data science algorithms, the future is conversation. Take Amazon’s voice assistant, Alexa for example. This is just the beginning of what is to come for voice assistants. They will soon predict what you want to do before you do it. It not only identifies which restaurants you’ve chosen in the past and serves those up to you when you’re searching for a place for dinner, but also understands your budget, and even predicts which cuisine you’re most likely to desire on certain days. In 2021, the focus is on recalibrating the models to make sure they become far more accurate and responsive, so they are able to predict more important things as well, like another pandemic in our future. These are the types of changes we are expected to see coming shortly.
Has COVID caused digital transformation to go on steroids?
Absolutely. Customers want the best service, and this is now, primarily digital. Companies need to ask themselves, “How do I make the customer’s experience more interactive?”. Things are becoming easier through this digital transformation, but we haven’t hit a breakthrough point yet.
COVID has affected every industry, but two that have been able to benefit from this digital switch is internet-based retail (such as Amazon) and the financial industry. For retail, foot traffic was used to determine if sales were likely to go up or down – now, this has switched to measuring website traffic and digital sales. For the financial industry, if you know what the sales are for, say, Nike, then you are able to trade this stock confidently. In the past, financial analysts had to wait for Nike to release foot traffic metrics at their stores in order to predict Nike’s future sales. However, today data scientists can analyze cell phone signals available from telephone companies and cross link number of phones in front of a Nike to store to predict future sales. They do not have to wait for Nike and thus trade its stock ahead of other financial analysts. Data science has made it easy for companies to be able to gather their own data quickly.
Additionally, there are models that analyze Twitter feeds to make these kinds of decisions. For example, if people are talking positively about a company, your company may want to buy that stock, and vice versa if they are talking negatively. More companies will begin using these types of models moving forward to make these types of business decisions.
Is data science something that only large companies get to take advantage of today?
Data science is available to both large and small companies. Today, data science tools are available for free, so all companies have this advantage. However, smaller companies have the disadvantage of limited resources and time. Data takes time to clean, transform, and productionalize. You must constantly update the data, as it is always changing, thus having to clean it, and update your model constantly. Smaller companies don’t have the time for this, so are unable to do this themselves. Additionally, there isn’t enough talent that understands the cloud, data science, and the business to know what is needed to improve the business. So, while the models exist, the tools to use data science and use it productively are not quite there yet for small and mid-market companies because they don’t know how to use the practical application.
During 2021, we expect to see a SaaS (Software as a Service) model for data science. Big companies are developing all data science models and associated infrastructure and offering it to smaller companies for a low monthly subscription price. Smaller companies will benefit from the advances of these subscription model services because they will not. to spend time and resources to develop their own software or data science models. Additionally, as times goes on, the expectation is that the price of these services will come down and be more affordable for the small- and mid-market companies allowing them to receive these sophisticated models at their fingertips to stay relevant and up to date.
Big Tech, the most dominant of companies in the information technology industry, are driving this technology forward to make it accessible to everyone, thus expanding its commercialization. They are developing models that are creating products more and more personalized to the consumer. These models are created to understand what a consumer wants, and when and how they want it. The more time consumers interact with these products, the more they learn about them, thus, the more personalized they become.
What do you recommend goes into a company’s data science strategy?
The main goal of any data science strategy should be to directly or indirectly increase revenues and profits for the company. Thus, chief data scientist must align data science strategy with the business goals – ask yourself, “how do we solve this problem so the business benefits?”. It must have a cloud strategy too, as every business is moving towards the cloud due to the pandemic. Identify the problems your company is facing and solve these problems accordingly using the best in class data science and cloud tools available. Rent vs building/buying these tools!
Is there anything else you’d like people to know about data science?
Error space sometimes kills you. The most important aspect of any model is knowing if the error is catastrophic or not. Just having a model is not enough, you must weigh the error space in order to identify if this model will benefit your company, and at what risk.
One thing to remember is that data science is used to help both businesses and consumers. It benefits both big and small companies, and its goal is to provide value and solve problems. Data science knows no bounds, and the sky is the limit for what is to come in 2021 and beyond!
About Surya Systems
Surya Systems has been a leader in IT staffing for over 20 years, serving mid-sized and Fortune 500 companies and their technology, and is one of the largest diversity staffing firms in the industry. We are known for our high-touch, customer-centric approach, offering our clients unmatched quality, responsiveness, and flexibility. We are appreciated by our clients for our streamlined execution, highly efficient service and exceptional talent management that go above and beyond traditional staffing services. Surya provides solutions to the airline, banking, energy and utilities, health care, manufacturing, telecommunications, and insurance sectors.