Nirupama:Hi, Shridhar. Thank you so much for joining Coefficient! This is the sixth episode of Coefficient and we are talking to you today about Connected Intelligence, which is such an important topic and such an interesting concept that we’ve been talking about a lot in MathCo. And, you know, this is an episode that I’ve been really looking forward to having.

My first question for you is very basic—to set things up. Can you tell me very briefly what is Connected Intelligence and why it is crucial?

Shridhar: Hi, Nirupama, really awesome being here today to talk about this topic. This topic is very close to my heart, so definitely, I think, this has been coming. I think, if I had to give an overview of Connected Intelligence, I think… Connected Intelligence comes across as a very generic term, right? So let me try and break down how we are thinking about it.

So most organizations today, in earnest, are trying to get to data-driven decision making. I think they are in their journey; they have potentially invested in a lot of tools. These tools help them store data, take some decisions—maybe even strategize and plan, right? So they’ve been using these tools and they’ve created all these bunch of assets within their organization.

Now, our aim simply, with Connected Intelligence, is to connect all these assets. And we want to connect them in multiple ways. We want to connect them vertically, so that they’re aligned to the top-down organizational objectives. And we also want to connect these assets horizontally, so that they’re re-used and people are not reinventing the wheel across different business units.

So that’s a very simple overview. Now, when we deep dive into this, I think you will see this expand for various different personas in different ways, but hopefully that gives you a sense of how we are approaching Connected Intelligence.

Nirupama: Yeah, it does. Thank you. And my next question was going to be about that. If you could maybe paint a picture of how it plays out and how this concept of Connected Intelligence manifests in so many different things that a company does?

Shridhar: I think a lot of such data-driven decisions take place in a lot of organizations; especially organizations which we partner with. So, we are partnering with a lot of Fortune 500 organizations across different industries, right? Let me try and take an example from one of those. Let’s consider a consumer-packaged goods company, right? So potentially they are trying to make sure their products make it into the market.

Now something which they do—year in, year out—is forecast demand for their product itself, how much of their product is going to actually sell in the market. Now, various drivers make it into this forecast. These drivers could be: what is the price of the product? What sort of promotions are running and so many more, right?

And each one of these drivers are owned by different stakeholders, right? So that happens in every organization. So how can we connect that and bring that together? Now, once these forecasts are ready, based on these forecasts, there are so many downstream business processes which get affected. I have to potentially do my business target planning based on these forecasts. I have to potentially plan my inventory, how my products are stored, how much inventory do I need to maintain? And finally, I think while solving all these point problems, they sort of lose touch with What is my organization’s business goal? What am I trying to run after, right? So if I do these forecasts, how can I plug them into business processes? And how can I track progress towards organization goals? So that’s how I would see this manifesting in an example, for a consumer packaged goods company.

Nirupama: Right. Just to follow up from that, I’m curious to understand, is this just an example or does this already exist in some companies? Or is this like a 2024 plan for a lot of companies?

Shridhar: No, no, this is valid and a very real example. And I’m sure folks who are either in CPG or in different areas as well, right? We are seeing healthcare companies do a rare disease forecasting. We are seeing companies do inventory forecasting in retail as well. So, I think this hits the nail for so many business processes, which folks are trying to pursue.

Nirupama: Got it. You’re the CTO and you lead the engineering function here at MathCo. I want to understand from your perspective, how Connected Intelligence affects data engineering and what is the whole technical set up behind the whole thing?

Shridhar: So I’ll add a couple of things here. I think one part is probably, uh, more experiential for me. Uh, so I come from a more technical background. I come from product companies being a software engineer myself. And this is the advice I give to anybody who works in engineering. Hey, you need to get your work closer to business.

So you need to understand what business process your engineering work is connecting into and what sort of impact it’s providing. This is the aim from Connected Intelligence as well. If we are creating data assets which are going to power the decision making, if we are sort of creating AI/ML assets which are going to power the decision making, we want to make sure—through engineering—we are connecting these to the business process itself. I think that is how at least we are looking at it. We want to make sure when one, engineers are working on this, they are aware of what they’re working through to and secondly, for our organizations (our partners)—when the end users use something, if they are looking at an ML model, I don’t want them to sort of have full knowledge of data science and engineering. If they know what business process they are of plugged into, they can make sense of this model and take it forward. I think that’s the thinking behind how I’m connecting this to engineering.

And secondly, I think we feel that sometimes, these foundations are very siloed in organizations, whether it is the central IT team, whether it’s the central analytics team, whether it’s the data team, or sometimes the business IT team, sometimes the ownership is spread across. But with the approach of Connected Intelligence, irrespective of this distributed ownership, we can at least approach connecting some of these assets towards that decision making.

Nirupama: So now we’ve gotten a sense of what Connected Intelligence means and how can we see this working within an organization and a little bit about the technical aspect of things, like what goes behind making this a reality from an engineering perspective.
So, maybe taking a few steps back, as a normal business user, how does this affect me if I’m a business user in a company?

Shridhar: That’s a great question. And, we get asked this question day in, day out, right? I think you come up with sort of an Uber vision like Connected Intelligence and business users are like, Hey, how is this changing my day to day?

So we tried our best to simplify this and make sure it resonates with the business end users itself. So number one, it helps them understand what organizational objective and initiative they’re working towards. All their efforts—what is it really feeding and what is the progress? Can I track the progress? That’s one.

The second one is because this is a world of multiple stakeholders and multiple decision-making, help them understand which stakeholders made the upstream decisions when the decision came to their hand? What is the decision they need to take and what downstream processes does this affect?

So that’s point number two. And point number three is, they have all these solutions in front of them: a bunch of reports, a bunch of simulators, bunch of Excels, word documents; How to make sense of these solutions, so that they can do two things: answer their questions, and take their decisions, right?

So I think that’s the idea: really connect this to the organization; connect this to the stakeholders and help them make sense of the solutions and take those decisions. So that’s how this is going to help the business end user in their day to day.

Nirupama: Okay, so I don’t know if my understanding is correct. You can correct me if I’m wrong. So, will this make it a lot more transparent and hence, not just, you know, a lot of opaque decision making and top-down approach to a lot of decision making. Will it increase the possibility of people from different departments and people from different rungs of the ladder sort of giving feedback and sharing something that might help fix a mistake or maybe find a better way to do something that somebody in the other department might have not noticed?

Shridhar: Absolutely. I mean, you hit it on the head. I think through first principle thinking also, I mean, if you connect these things, you can see that transparency will be increased. But, we should not look at this as, Hey, there’s the silver bullet solution; you set up Connected Intelligence and day two, it’s all transparent, right? I think it has to be done deliberately. So you need to meaningfully connect your goals to your business processes. It should not be transparency for the sake of transparency, right? So the idea is, not just connect it to business process, but to the goals as well. So you get to see what progress the organization is making.

Nirupama: Right. So, following up from everything we’ve discussed now, I’m curious, you know, today we are talking about Generative AI, and in a lot of our solutions, we are now using generative AI and we are at the forefront of that. How does generative AI play into Connected Intelligence?

Shridhar: Oh, that’s a great question. And I’m sure everybody is thinking about generative AI right now. It’s all over the news. Um, I think one thing I wanted to talk about, I think, when we think about Connected Intelligence, bringing together your own assets, talking about owning your intelligence, right? Artificial Intelligence is a big part of that. So, with Connected Intelligence, one of the aims is to unlock the potential of analytics and all these new technologies. So if we are using artificial intelligence, and I think artificial intelligence is this umbrella term, which consists of Gen AI, which also consists of your existing AI/ML models. So, one is we want to make sure there is that Connected Intelligence space. So, you can bring in technologies like Gen AI, and if they generate some intelligence, it can be plugged into your day-to-day processes. So I think that’s one. Second one is, since, in the Connected Intelligence sphere, we are really looking at decision making, it forces us to make sense of Gen AI technologies. So we are really working on some cool uses of Gen AI technology to help the decision making. And again, I’ll go back to those two steps. It helps us sort of summarize existing assets. And answer questions and eventually they help me take a decision in the form of copilots and sort of similar agents or bots, but it is truly relevant with this space because it has the word intelligence built into the topic.

Nirupama: Yeah, it connects back to, a couple of episodes back we did an episode on Gen AI and now I see the, I connect the dots between, you know, why we’re talking about that and why we’re talking about this.

Now I want to break this down a little bit more. So we are talking about Connected Intelligence, and I understand that it is something that is very beneficial. But, is this for all companies? And also the companies that do want to achieve this, where do they start? And also who would own this mandate within a company? Yeah, so these things… like just like breaking things down a little bit more.

Shridhar: Sure, absolutely. And I totally understand because we are talking about such an Uber theme as Connected Intelligence. It’s difficult to envision, hey, who is going to get started with this? And does it really need to come top down? Right? We have worked with all different forms of organizations in this endeavour, and we feel there are multiple different starting parts.

It’s not a one size fits all. So, let me take a few examples in that nature, right? So, let’s take the example of executive business users. They are trying to strategize and potentially plan for the next year or so. So here our aim is to connect their strategic business objectives and goals to the business drivers itself or to the workflows and then to track the progress.

And that’s it. So that’s the starting point with, if you will, business executive users. Now let’s come to business users who are potentially taking day to day decisions. They are measuring the impact of these decisions and affecting change on the ground, right? So, for them, we can just connect their business workflows to their solutions.

And that’s a small enough starting point for us to get started with business end users, right? And similarly, if I look at the central data or the central analytics units, we can think about connecting the data or the AI/ML assets to the business workflows, right? So that’s the idea.

You can now imagine that there are several starting points to this puzzle and it’s not a one size fits all. It’s not a one project sort of a journey, but it’s a journey towards maturity and connecting various different levels of assets here.

Nirupama: Right. Again, I’m making connection. Just the previous episode we did on revenue growth management and we spoke to Silvana and Aditya and they were talking about how in the maturity curve for revenue growth management, the ideal place—you know, high end of the maturity curve—is the integrated model.

And you know, that again comes back to this, right? Where you connect your decision making that is focused on one thing like revenue growth management connected to other things within the company. So yeah, that makes a lot of sense.

So, we are talking about connecting so many things in the backend, and I’m sure there are already so many tools involved and, is this truly going to simplify, you know, processes and decision making for companies and leaders? As in, how do I know that this isn’t just one other thing that I’m adding on to everything else that we are already doing and the tools, the several tools that already exist?

Shridhar: So, right off the bat, right? We are not trying to introduce a new platform or a tool into this space. We totally understand that in this world, whether I say a data platform or an analytics platform, there are so many tools which are lurking around and I’m sure organizations have invested in them as well.

So this is more, um, I would say an approach towards data-driven decision making and an approach towards connecting existing assets with open-source technologies. So, what we bring to the table is a modular approach with these open-source tools or modules. We bring them into your ecosystem.

Now, you might either have procured a license for existing tools, you depend on a vendor to build these. So the idea is to bring these tools into your ecosystem, help you build your intelligence assets and own this intelligence, I think is the end of the deal. Instead of we trying to push another platform or a tool in this space.

Now I think this has manifested in different ways in some organizations. This leads to de-duplication of work. We have seen that, for the same solution, multiple tools are being used. You end up realizing, Hey, I’m using multiple tools in this space. Maybe I can cut some of these in some space, in some organizations.

We have seen that some of this work has traditionally been outsourced and the vendor owns the intelligence rather the organization itself. So, to bring that in, you would have to bring it into your firewall. A few open source tools so that you can own and maintain this. I think that’s how it manifests.

And the idea is not to introduce new tools, but to make sure existing tools and assets are sort of brought together and connected.

Nirupama: Yeah, that brings us to the last question that I had, which is, you know, how important Connected Intelligence is to MathCo. How it is like sort of foundational to pretty much everything we do here.

And in fact, our new tagline is Own Your Intelligence. And I think you said that a few times in your answer. So, can you talk a little bit about how this ties into everything we do at MathCo and also, you know, a little bit about how it was discussed for the first time, how did it become this thing that sort of anchors most of the things that we do here?

Shridhar: Yeah, absolutely. I think, as we grew in our journey, I think, we have been solving a lot of point business problems. Then we started having longer relationships with our clients. We started becoming their strategic partner, their preferred partner, helping them make some strategic decisions.

So, what we saw was: One, the experience we were delivering was across the analytics life cycle, whether it was on the data engineering front on the data science front or the decision-making front. The second thing we saw was, we were working with various business units for an organization and they were potentially running a different way. They were using different tools in certain places. They were using different clouds even and the idea was: Hey, I think if we were we are solving problems like this, We have become a strategic partner for our clients, Why not look at how can we minimize some of the duplicate work which is happening here? Make some of the decision making easy and also create a space where all these new technologies can come in and fast track or accelerate some of the process towards maturity, right?

I think that is what gave birth to Connected Intelligence. I think we have been potentially doing these items without naming it Connected Intelligence for the good part of last three years. But I think we have been able to put a brand around it and very clearly articulate How this helps an organization and what sort of journey can they go on and how can they see impact here.

Nirupama: Thank you so much Sridhar. For closing remarks, if you were to say, you know, in 2024, where do you see Connected Intelligence; companies progressing in terms of Connected Intelligence, both in respect to our clients and maybe even outside of that, what would that be like?

What is your vision for 2024?

Shridhar: Yeah, I think, as part of your research itself, right? We heard a couple of Gartner analysts, speak about decision intelligence, and I think they are speaking about one, the focus on cataloguing decisions and utilizing these decisions in the future so that they can make better decisions, right?

Definitely I see a lot of that happening as we go forth, because right now decisions are scattered across various tools. How can we bring them and connect them. That’s one. And two is, I think you have seen, the advent of Gen AI across this year, and I think it’s only going to improve from here on.

So, how can we make this space so that, as Gen AI improves, we are modularly able to plug in new and newer technology pieces and accelerate our decision making here? So I do see a lot of potential and opportunities in this area. I also see that for executive users, I think traditionally there is a lot of manual reporting, where a lot of that is going to get replaced with interactive tools, which is also going to make sure a lot of these intelligence is democratized. So I do see a lot of growth in this area, and I’m really hoping that NucliOS, which is our foundation of Connected Intelligence, is able to provide some of these modular aspects to aid and accelerate this journey for our partners.

Nirupama: Thank you so much for taking the time to talk to us about Connected Intelligence. It sounds very promising and like you said, it’s not a silver bullet solution. And we need to think about it carefully and implement it carefully. Thank you so much for joining us today, Sridhar.

Shridhar: Thanks Nirupama.