Tellius’ CEO Ajay Khanna: Closing The Data Insights Gap

Tellius is an AI-driven decision intelligence software platform company, backed by Sands Capital Ventures, Grotech Ventures and Veraz Investments.

I spoke with founder and CEO Ajay Khanna about the inception of their core idea – combining AI and machine learning driven automation with a natural language search interface for ad hoc exploration – and leading as a technical founder.

Karen Walker: To understand Tellius, look at the industry?

Ajay Khanna: There are two silos in the industry. On the business intelligence side, great companies like Tableau and Snowflake do an excellent job of creating visualization and enabling an organization to do the reporting. But most  companies generating large amounts of data want to understand both what happened and why things changed.

We saw that the current business intelligence didn’t address the entire issue because it will give you reports, but you still have to figure out why.

Now, on the other side of the silo, there are machine learning and AI tools coming from some great companies, older companies such as IBM and SAS and new companies like DataRobot. They are doing an excellent job of creating machine learning tools and capabilities.

Most of those are still catering to the individuals who understand data science and machine learning techniques. So when we look at the business analysts and business users who want to leverage machine learning and AI to make faster decisions, they can’t leverage it unless they have a team of data scientists.

That’s creating what we call a tremendous insights gap in the organizations – meaning they have the data, storing some of it in the cloud, but they don’t generate insights from there.

We set out to resolve or bridge that gap. In 2016 we started Tellius, and our platform combines the ease of use of a natural language interface with the power of machine learning and AI.  

You can ask a question, “Show me the revenue by region and product.” And the system will then word that into a SQL query, go to the data warehouse, and that could be terabytes and terabytes of data, and get the answer in a matter of seconds.

That’s our mission – to democratize data, to use machine learning and AI to enable that experience, and get those deep insights from the data.

KW: Can you give me an example?

AK: It’s not necessarily the complex problems – those exist and do need to be solved. But there is a lot of need for support with the day-to-day operational issues. After COVID-19 hit, the kind of information people had and what they used to make decisions completely changed overnight. 

A recent example comes from one of our pharmaceutical life sciences customers. Their question was, “Why did the market share for a particular drug go down?” They had a lot of factors to analyze: Are our other drugs co-competing? Are the providers not prescribing? Is it a demographic change? They also had preconceived notions about targeting tier one healthcare providers.

With the click of a button, the system was able to zero in on the problem – and it wasn’t in the leading tier one providers that they had expected. It was in tier two and tier three. Their competitors were targeting those tiers and gaining market share. So they changed their strategy, particularly around education, and were able to increase their market share.

KW: So the company is uncovering blind spots with data. How do you do that?

AK: Machine learning is just classification – “good time” segments when things are working and “bad time” segments when they aren’t.  For some of these questions, people go to data scientists and spend a few days or weeks to get the answers. That  is a slow process, and you can only answer a few questions, depending on the data scientists.

Tellius enabled the ability to democratize this – you click a button in your BI dashboard and drill down from there. That button triggers machine learning algorithms – finding the difference between the “good time” and the “bad time” and then surfacing that to the user. The analyst, who understands the business, can discard irrelevant or unchangeable factors, and then they iterate. After a few iterations, you can get to the answer, which is a good combination of human domain expertise and what the machine can show you.

KW: In your bio, you declare a passion for delivering a fantastic customer experience.  What does that mean to you?

AK: We must answer three questions to deliver that experience: How do we reduce friction? How can our product help customers get insights from the data? How do we enable them to leverage the technology?

KW: You have a deep technical background, double engineering degrees.  Can you describe your first leadership experience?

AK: It’s tricky. I certainly had a learning curve, and I’m still learning. I’m focused on enabling everyone to rally around a mission. That’s one of the most important things when you are a young company in a field of big players trying to change the world. Because with a small number of people, everyone needs to be a leader, in alignment with the mission and empowered to make things happen.

KW: On a 2×2 matrix, we always want to move up and to the right. Think about your career – when were the moments when you knew that you were moving in that direction?

AK: I learned that in order to succeed, you need to have four things: an agile mindset, ability to listen closely to the customer, a mission in which you strongly believe, and an agile product.

In a previous organization, we had a services company and then launched a product. Because we were listening closely to our customers, we realized that we were running these as separate organizations and – this was the pivotal “aha” moment – that combining these things would bridge a gap. Only because we were listening closely to the consumer, were we able to gain this insight. At Tellius, we are still in the very early stages of growth and the need is changing rapidly. We have to be agile and listen to the market to be able to move into that upper right quadrant.

Originally published in Forbes.

Share this article on...