We bring to you a 3 part interview series with the BRIDGEi2i Leadership. Third in the the series, we talk to Pritam Kanti Paul, Director and Co-Founder at BRIDGEi2i. In this interview Pritam talks about advancements in the field of Statistics and Machine Learning and where the Analytics is heading.
Here is the transcript of the interview:
INTERVIEWER: We would like to know more about how you started your career?
PRITAM: I am a Statistician and did my masters from Indian Statistical Institute. I moved from there to GE when it was starting its Analytics Centre of Excellence in India. It was a good opportunity to work in an unstructured, growing environment. I started with consumer finance and worked with various countries and functions in GE and got a global exposure. I also tried my hand with one of the Start-ups at that time which was trying to integrate analytics into Call Centre operations.
In GE, I got an opportunity to set up an analytics consulting team supporting strategic projects across all the businesses of GE. When GE became GENPACT, we also got the opportunity to work with some of the pilot customers at that time. And then I moved to HP and it was a roller coaster ride – starting from scratch to build the team, telling them what they should do in Analytics and another five-six years went by a whisk where we created a thousand plus member team working across all the different businesses of HP. And then our dream company started which is BRIDGEi2i and the excitement continues.
INTERVIEWER: Statistics, per se, is a very vast field. So how do you see this field has changed since the day you started to present day?
PRITAM: Today it’s a different world since there has been significant amount of improvement and advancement in the area of Statistics. Obviously, academicians can tell much more about it. In my mind, from the day I started in 2002 to now, there has been a significant change in a few parameters in how Statistics is being used today in business problem solving. One is expectation of speed. We used to take two months to build such models and probably today people expect them to be built in a week or two. As people say analytics is about exploration and if you are failing you should fail fast and hence speed is critical.
Second there is a significant amount of awareness and excitement today created by use of analytics in different fields. As I talk more to my customers today I see lot more excitement about how analytics and statistics can make difference to their business. A lot more people want to use it, they might not know how to use it and it is our job to make it available to them. And on the other side, because of advancement of techniques and because technology is able to handle those techniques the opportunity of using analytics in certain areas has significantly improved.
INTERVIEWER: Another important buzz word that we are hearing now is Machine Learning. So the Analytics Industry is using Statistics and Machine Learning both hand in hand. Please share more on this.
PRITAM: That’s where some of the parameters I talked about like speed and flexibility comes. Machine Learning is not new. Computer Scientists have been using for ages in pattern detection. The excitement in recent days is because of the need of speed and probably because of flexibility needed for somebody who doesn’t know analytics should be able to run it in some sense or the other. There is a need for self-learning from the data i.e. automated knowledge discovery from the data. In Surveyi2i, one of our flagship products, there are a lot of places where we have leveraged it to make knowledge discovery very fast. On the other side, for example we work with customers who need online recommendations as to when the visitor is really visiting their website and it cannot be done by static models because it has to self- learn very fast and immediately give a recommendation.
That’s where traditional methods like statistics might or might not hold. Having said that, we should not stretch it too far to say Machine Learning is the end game and statistics is not the right means. I think there are different ways to solve business problem by data. Some of them where speed is required, Machine Learning is useful. There are other problems where we need to give much more importance to accuracy because not being correct might have huge penalty associated with it and that’s where Statistics in its true usual form comes into play. But I see a lot of opportunities of marrying the two in some places to gain maximum. It is not going to one or the other but the two of them together.
INTERVIEWER: So in a way you are saying that Machine Learning could replace Domain Expertise?
PRITAM: I am a strong believer of analytics being a conglomeration of both statistics and domain knowledge. Many of the times analysts think that most complex and challenging problem to solve is where they can use the more sophisticated technique than just statistics. But we should think from the customer standpoint as in what is the most important problem he is facing at that point of time. The analytics techniques should be used as per the demand of the problem. And as I said, if you need a very fast and self-learning solution we might need to go to Machine Learning but it doesn’t really take out the need of thinking through the problem and come up with a structure on how to solve the business problem.
One area where machine learning sometimes helps is what you call as Knowledge Discovery. There are cases where we start with a hypothesis about the problem statement and seek that knowledge to solve the problem. In some cases Customers might come with the Data not knowing what the problem is and want to go for an exploration. That is where machine learning becomes handy to come up with various potential hypotheses. Once done, we need to marry this with domain knowledge to make sense out of it. So what might happen is for people who are strong in domain knowledge and not very strong on analytical side, we might be able to give them certain tools and apps like Surveyi2i which can probably enable them to run Analytics.
INTERVIEWER: While we are talking about Machine Learning, another buzz word that is making news is Unstructured Data. This is a monster the Industry feels, is very difficult to tame. How do you go about it?
PRITAM: These are all parts of the problem that we are trying to solve. Unstructured Data was there for all the while. Now because of technology enhancement we are able to put the data in place and make it accessible. There were challenges in terms of cost, speed, to be able to put it into a place and make it accessible to people. Today’s technology has made that accessibility much easier. We have many social media listening companies who are tracking all unstructured data and putting it into a place because of low storage costs.
The challenge lies in how to make sense of the data and probably decide what is important from this huge sea of information. Textual information is there in several different places. How to create indices from social media and identify how those indices have correlation with some of the business metrics is important. BRIDGEi2i works on figuring out some of those. One of the projects we did was to predict new product success and we correlated sales with some social media buss and quality indices created by us. The challenge lies NOT in techniques and technologies since a lot of them are now available but in making sense of it and going beyond the obvious and usual. I think the industry has probably moved a little bit into that trap. Everybody is talking about sentiment and word cloud. We need to go beyond that and that’s what the industry is working on.
INTERVIEWER: Continuing with your thoughts, from the Industry standpoint, they are apprehensive in big data and unstructured data mainly because of deficiency of proper resources, systems and processes. How do you see in the future it will evolve?
PRITAM: The positive of big data and unstructured data is that, it has created a lot of excitement. The positive aspect of it is awareness and negative aspect of it is potential diversion of focus from impact. I see two different problems happening. One is your business situation might need you to analyse your internal data which might not be overtly great at first and there may be lot more opportunity there and you still going behind the so called exciting area.
Other area is that not too many companies have managed to store and make the data accessible in a simpler way. I think if the Industry has to really sustain the momentum on big data analytics in the next couple of years, we all need to start thinking beyond storing data and running mathematical queries. We have to really think about what is the business problem that needs to be solved. Be it small data, be it big data, it does not matter. As long as you are making impact to the business there is a significant opportunity.
INTERVIEWER: Talking about big data, is it possible to create a mobile app using this big data?
When you are solving a business problem one should not always restrain themselves in simplicity of method. Complexity is good for people who know how to handle that complexity. So as an analyst, as a Business Analytics company, I will be trying to find out the best way to solve that problem. There are significant number of people who are going to consume such analysis and they want to access analytics in a much simpler way. Now that is where the complexity has to be taken out in front of those set of savvy business managers who can contextualise with the business problem, look through numbers, be able to simulate situations and take a call. But the front end on which somebody is working on has to be made very simplified. That is where BRIDGEi2i focusses a lot where we are trying to build certain platforms and algorithms which try to solve certain set of business problems and create a navigation which might be mobile enabled or web enabled to end users.
We have our flagship product Surveyi2i which basically allows an Analyst to quickly analyse survey data. For the analyst, it is about the speed and being able to do certain things very fast and allows them to run it by themselves without going into the technicalities. Similarly there are other products which we are building in certain areas which will come up in the next one year or two with similar kind of features. Front end being very simple, back end being probably quite complex.
However I don’t relate complexity with the value because there can be lot of complexities to make sense of the big data, get them into so called few numbers which somebody can digest and provide a simulation engine where people can play around. So I absolutely believe there are lot of opportunities to create some of those navigations and make it much faster.
It is important to create a bigger eco system. As a Statistician we believe that this is my job and after that I’ll send some numbers and somebody will take a call on the other side. They might not even know what is happening. I think if you really want to make analytics as the next big integrated function it is important to include others in the process and I feel that some of those mobile applications which you are talking about will include lot more number of people who are front end managers in the process of consuming analytics.
INTERVIEWER: Pritam, there are already a lot of companies in this big data field. Is BRIDGEi2i late entrant in this industry?
PRITAM: With the big data wave there is huge amount of opportunity for any business today. I don’t think there is an issue of a very concentrated market where Bridgei2i will have problem. For the last few years we have fared very well. Having said that I think the nature of the problems are changing significantly in terms of speed, flexibility, need of being able to quickly solve certain business problems, solving veracity problems.
What I am trying to say is there is even newer fields and new way of solving problems. Some of our very large competitors, they might be solving certain business problems but the nature of the problems might change tomorrow. From that standpoint we are not late entrants. And I also don’t see too many companies who are talking in the language of BRIDGEi2i – namely being able to solve the business problem, being able to integrate technology solutions where we are creating a space for ourselves. And I have not seen a huge competition at this point for BRIDGEi2i. I don’t also see some of the technology companies being a big competitor of us. I see them as enablers where we use some of their web platforms and others. I see a lot of opportunities for us.
INTERVIEWER: We would like to know more on what BRIDGEi2i is working on?
PRITAM: As an organisation, we have chosen a few areas like Customer Experience, Customer Intelligence and Risk. We have identified opportunities of creating algorithms. I think you asked that question some time back about Machine Learning. I strongly believe there are areas where you can use it. We have taken few of those areas and we want to build technology platforms on top of that. Some of those platforms can be taken to a shape of a DIY application like Surveyi2i.
Some of our products are just platforms which need to be customised on for a specific enterprise. Survey Analysis, Customer Experience and compliance are some big areas where we are working in. We are also working on Risk Governance as huge area. It is a big area for Financial Services where they want to track a lot of different models across the world. It is a platform which will enable everyone across the organisation to get a single view of all of them together. And we are also working on whole area of micro-segmentation. These are the top three areas, I would say.
Analytics is all about exploration. We are exploring several areas where big data can be really integrated with some of such contexts and I think some of them would evolve as we move forward. Probably next time when we talk, we can talk a few others.
PRITAM: It was a phenomenal experience. We saw a lot of different products which were being showcased solving very core and futuristic issues in Financial Services. To be a part of itself is a matter of pride and excitement. We met a lot of people, we understood where the industry is going but the largest earning from the individual perspective is that it was basically a seven minute stage presentation without a single slide and we had to demo it.
It typically takes forty five minutes to know the product and we had to talk concise in seven minutes. It really creates a lot of clarity in mind about what is important about the product. It’s a great learning on how to present your work in a few minutes. That’s the most important learning from the individual side. We also saw the crowd appreciating the product which gives us the feeling that we are in the right space and the product has a lot of opportunity. So, Great Excitement and a great place to be in!
INTERVIEWER: And last but not the least, what would you suggest to people who want to make it big in this Analytics industry?
PRITAM: I would say for the individuals, there is a lot of excitement in this area. For the industry, it will significantly grow. We need people in this industry from all the different parts of the world as well as different types of disciplines to come together. But at the same time, remember if you are really excited about solving the business problem, about the mathematical problem, about analytical thought process then join this industry. Don’t join because there is a huge amount of buzz around. Unless and until you like it, it is not an area where you can grow significantly as an individual.
From a company standpoint, I am a strong believer of the fact that companies use Analytics to create an impact and reinvest that money into Analytics. There are only a few companies who are Analytical competitors. They use analytics as a part of business processes but rest will look at Analytics as a creation of Business impact and if that really happens they will make more and more opportunities. My suggestion is to obviously go beyond continuing to do what you are doing and look for where you can make an impact for the business.
BRIDGEi2i provides Business Analytics Solutions to enterprises globally, enabling them to achieve accelerated business impact harnessing the power of data. Our analytics services and technology solutions enable business managers to consume more meaningful information from big data, generate actionable insights from complex business problems and make data driven decisions across pan-enterprise processes to create sustainable business impact. To know more visit www.bridgei2i.com
The views and opinions expressed in this article are those of the author and do not necessarily reflect the official position or viewpoint of BRIDGEi2i.