Skip to main content

Avijit Datta is CEO of Aviana Global Technologies, a company dedicated to working with organizations that need predictive analytics to make better decisions, effectively manage risk and improve profitability.


How did you get started?

Back in 2001, I decided to start a company specializing in custom application development. My partner, Ananta, decided over a glass of California Chardonnay at a restaurant on the Pacific Coast Highway, to name the company Aviana, which is really the first part of my name and the last of his!

For the first year, we did a wide spectrum of application development work. However, we soon decided that we did not wish to be everything to everyone and that we had to specialize in something that used our deep knowledge in data manipulations and management that we had gleaned over the years. Around this time, the emergence of Business Intelligence was seen as the next biggest step toward using technology to make sound business decisions. The shift to business intelligence was a natural direction for us and an area where we had an incredible number of success stories to tell. We decided to closely align with Cognos, and then IBM acquired Cognos, making us an IBM partner. For the next several years, we were the go-to partner for the West Coast.


Getting Data Right Becomes Critical

While reporting is a critical piece in effective decision making, we realized that if we don’t have the right data in place, we are not ready for Business Intelligence. Any reporting from this data would be erroneous and misleading. We realized that the level of ‘data maturity’ varied across organizations and not every organization had the skill, willingness or the person power to get them there. That’s how we got to be very proficient in Data Warehousing. We had to take the responsibility of getting the data ready for meaningful analysis.

Successfully dealing with massive and complex data for business leaders across a number of industries gave us the depth of experience that rocketed our reputation nationally. We were there at the very beginning of the data revolution and have not faltered in our objective of being the primary data management organization in the business.

Somewhere along we got into a partnership with SPSS, which was also acquired by IBM later on. At that time, SPSS was into basic predictive analytics. It was a little ahead of its time because people were still getting used to the whole business intelligence concept but it did allow us to get our ‘feet wet’ in the predictive space.


Expand into Data Analytics

In 2016, with the sole objective of getting into advanced analytics like predictive analytics and anomaly detection, I bought a company called Elite Analytics, a Texas-based predictive analytics company that had done some great work, especially with governments across a number of different states. Predictive analytics is a logical next step for those who had a strong Business Intelligence practice.

Business intelligence is all about post-mortem analytics and looking at data in the past, while predictive analytics is looking into the future and making educated and insightful predictions.

Detecting Fraud

Elite Analytics had done some excellent work in anomaly detection to identify fraud tax preparers for the California Franchise Tax Board in addition to a number of other government outfits – such as New York, Quebec in Canada, and a few others

Application of anomaly detection technology is a very industry-specific and complicated process and, consequently, expertise in it is rare and very much in demand. One of the foremost uses of anomaly detection is fraud detection. Aviana’s expertise in fraud detection encompassed tax agencies, welfare organizations, online testing companies, and automobile manufacturers. Aviana’s predictive capabilities were used by aircraft manufacturers to streamline their assembly process, by hardware manufacturers to detect overseas grey markets, and by recruiting agencies to find candidates that best fit their hiring needs.


The Birth of NEMESIS

In 2018, we decided to build a product to capitalize on the knowledge and technology that we had mastered over the years. Hence, NEMESISwhich is an advanced analytics platform that allows managers with no technical knowledge to build their own predictive reports at a fraction of the time it took in traditional organizations. It put the control into the hands of the executive, and dramatically enhanced their ability to respond to changing business circumstances.


What have been the most meaningful initiatives you’ve worked on?

There have been several. Streamlining the production of Boeing’s Dreamliner aircraft, helping DirecTV avoid a disaster with their then brand new black boxes, and helping iHeart media maximize their customer satisfaction metrics amongst a host of other such business applications have made this path worth every ounce of energy we have put in.

The ability to give executives or their teams the power to translate their knowledge and their expertise into actionable insights, and to help them to realize their potential by allowing them to use these tools that they never thought they could use by themselves, has been huge.

We have had some very touching experiences while some are very business-oriented experiences. It’s varied and that’s what gives me the most satisfaction. When you see the realization dawn on our clients who are using NEMESIS, that they now have the power to chart their own destiny and that we have forever changed the way they do business, it makes it very rewarding for us.

One of the rewards of using NEMESIS effectively is to bring out the potential of every team member in your organization. Their experience is valuable and the correct use of that knowledge helps the organization in ways in which they never could have without a platform like NEMESIS at their disposal.

For example, we are talking to a company that plans to use NEMESIS to help patients who have end-stage renal disease receive the best care they can afford. If we could help people have a smoother end-of-life experience, nothing matters more to us than that.

We are helping one of the largest online testing companies find fraud all over the world with our predictive analytics platform, NEMESIS.

We are also helping radio networks optimize their programming, helping credit unions maximize revenue out of their current membership base, and helping healthcare companies make sure that the efficiency of their medical testing is done correctly.


Most companies have some form of analytics. What’s different about predictive analytics versus traditional business intelligence?

It’s the difference between looking at the rearview mirror and looking ahead, or even beyond what you can see.

Good predictive analytics go beyond what you can see. It deciphers the signal from your data. It gives you the message that your data is telling you, which is not always obvious to the naked eye or just analysis.

Predictive analytics is what the executives would love to master. They always want to know what’s ahead, what if this happens, and what happens to my company, my profitability, or my longevity. Predictive analytics is where all the industries are moving towards.


What kind of results can executives expect from a well-implemented predictive analytics tool such as return on investment, reduction of execution risk or increases in profitability?

One thing that they can do, if properly implemented, is timely actionable insights.

Time is money. If they’re not timely, it makes no sense.

If they’re not actionable, it makes no sense.

It has to be timely so that they can prevent something from happening or fire off something to minimize the loss. Actionable insights need to be at the level of the user. If you give a manager a task that requires a vice president’s powers it is not going to happen or happen in a timely manner to be effective.

Timely, actionable, appropriate insight is the key.


It sounds like there’s a gap between strategic insights and tactical insights from predictive analytics.

Yes, there is. It is observed that whittling down strategic insight into tactical insights is not done properly in many organizations.

When implementing analytics in an organization, you start from the top and you whittle it down to to-do tasks for the people who are in the field. That’s how you get the cohesion. And that’s how you get that integrated approach to solving a problem and directing the company.


Many of the enterprise tools, such as SPSS produce strategic insights and they’re used by executives to direct companies. But there’s really no good tool to provide tactical insights. What do I do about it?

You break it down.

It’s like if you liken it to a person who has decided on the kind of person he wishes to be in the future. That’s a strategic direction. But what does that mean in terms of his habits, his values, his priorities, how he interacts, or how he thinks? Those are tactics.

In an organization, the CEO or the management decide what do they want the organization to be known as and what kind of reputation. Then break that down into doable actionable items that can be implemented by the levels below the management team.

That’s what NEMESIS can do because it forces you to think through all those levels of objectives. It is a cohesive integrated approach to solving a problem or creating a direction for the organization.


What questions must executives ask themselves when considering predictive analytics initiatives?

In a very broad sense, it is and I liken it again to a person is “who do you wish to be”? It is the most powerful question people can ask themselves – Who do you want to be?

An organization needs to ask the same question. Who do we want to be? We begin from there, and as I said earlier, we break it down into different levels of actionable tactics. Then create this cohesive approach.

So the questions are, do you know who we want to be? What do we need to know to be that, what steps do we need to take to be that, and what tactics do we follow to support that image that we aspire toward?


What questions must executives ask their team about a predictive analytics initiative?

What do you need to know to be more effective in your position? What information do you need to get? When and in what granularity? So that you can be more effective in your job position.


How can an executive best manage predictive analytics initiatives?

It starts from the ground. You have to have clean data and authentic information that is granular to the lowest level so that you can aggregate it based on whatever level you are catering to. You have to make sure that it’s free of any outliers so that you don’t get biased results.

Data is the lifeblood of an organization. Once you have the lifeblood, you know what to do with it. An organization doesn’t always realize this, but you cannot invest enough in making sure that you have the right data, right time, and the right granularity.


Then once they have that, what’s the next step to managing an initiative?

To understand what the objectives of the organization are. To understand what strategy you have decided to use to reach that objective. To break that strategy down into granular tactics and how that data will help you implement those tactics and therefore the strategy.


What big mistakes have you seen executives make when bringing on a predictive analytics initiative, and how can they avoid these mistakes?

Unclear objectives. They start at the tactics level. They have messy data. When they get the wrong results, they blame the solution. However, the biased result is actually caused by the messy data.

This is an iterative process. In predictive analytics, you do not hit the right answer the first time. But if you have the patience and the insight to see that you go down that path, the end of it is the bottom goal.


You’ve seen other analytics tools purchased but never implemented. Why does this happen?

Usually, that happens when the reasons for buying the tool are not in line with what NEMESIS brings to the table. When there is a problem that occurs, they want a quick fix. They haven’t thought through, or understand the strategy behind Nemesis.

It’s a very tactical tool that solves a problem. And yes, NEMESIS can solve that problem, but there’s so much more than nemesis can bring to the table, which is sometimes left out.

The successful implementation of NEMESIS requires a top-down approach to implementing it and visualizing how it can help grow the organization by empowering your analysts and managers.


Why do customers choose Nemesis even if they already have some sort of predictive analytics system?

Largely because NEMESIS was built for action. It was not built for only getting some analysis done. Anyone can understand their data with NEMESIS. They can choose from a spectrum of different predictive technologies and eventually get timely actionable insights.

The power of self-learning that NEMESIS has refines the insights to be better and better over time so that whenever you have new situations or new trends, NEMESIS absorbs that through the data and implements it through the engine.


How should an executive get started?

Right at the top. They have to ask themselves as I said, the billion-dollar question is – Who do you want to be? What do you want your organization to be? And then whittle that down to strategies and tactics and then implement it.

What are your closing words of advice?

There is a gap in most professionals’ minds about using technology and the knowledge of technology is beyond their grasp. And what NEMESIS does is to put it in their grasp, allowing you to have objectives and targets. That goes beyond your grasp so that you can be everything you’re supposed to be