09 - Reshape Business through Science and Data with Biju Domininc, Fractal.ai

Welcome to the certified skills and AI and tech broadcast brought to you by Criterion. Criterion is a global leader in assessment solutions, providing software for test development and delivery, as well as test development services for low to high-stake certification programs. So welcome everybody to another episode. Today we have Biju from Frachtle. He's a chief evangelist at Frachtle. And as always, we have Buzz Walker, our chief revenue officer here at Criterion. Welcome Biju, how's it going? Good, thank you. Glad to be talking to both of you today. Awesome. Well, let's get straight into it. We'd love to hear a little bit about what your company does and what you do with your company. Frachtle Analytics is one of the pioneers of data analytics and AI. It's a 21-year-old company. It's headquartered in India in Mumbai. But all our business is between US, Europe, Australia, and some parts of other parts of Asia. So that's where our business is from. As I said, we pioneered data analytics somewhere because it's about 21 years back we began. And today we believe that obviously the new wave of artificial intelligence is very strongly based on data. And so we are definitely taking advantage of that and then trying to work out the one of the leaders in artificial intelligence in the world. It has a definitely interesting point that you mentioned about data analytics. Everyone talks about data analytics and how it's important to have analytics for your company. How has AI in the last year shifted the way that you approach data analytics? We actually feed into-- I think what we have done is, yes, we said we began with data analytics. And but I think we also looked at certain new directions. And one of them was that we said that if we have to really look to the future and solve some of the most significant problems of our clients, we need three legs. And one is artificial intelligence, AI, then engineering, of course, to support that and design and behavioral science. And I became one of the part of fractal that brought in this new understanding of human behavior and design into fractal. And we believe it's this combination that is actually is going to make the solutions that we develop for our clients, palm-o, effective, and palm-o-palful than the traditional approach. What is a typical engagement with a client look like when you think about the human aspect, and you think the AI aspect, data analytics? How does that engagement typically look like with a client? It does not begin. The whole approach does not begin with AI. It actually begins with the client's problems. Or the client's problems, actually, if you go back, will always be about some decision that we need to take a deeper understanding of. So much so our deeper philosophy says, "Powered every human decision." So for us, it's all about understanding every human decision. And once we develop a deeper understanding of that, there has to be that decision has to either be made repetitive or there could be a new sort of a new decision or a new behavior that will have to be introduced. And that is where we actually bring in artificial intelligence. Because earlier, that might have been done by just human hands. And obviously, the efficiency would have been lower. But now, I think with artificial intelligence, we are able to make that process far more effective. - Very interesting, B.J.O. as you start looking at powering decisions. I also see that one of your platform solution areas is responsible AI. And wondered if you could talk more about that as we talk about powering decisions, all the different people that are making decisions, decisions to be made, data all over the place, lots of different ways you can use and misuse the data or the power of algorithms and what might be happening in AI. What is fractal doing on that part? - I think the very fact that we've fractal looked to really develop a deeper sort of an approach into responsible AI or ethical AI is actually is because we really truly believe, see the power of what AI can do. Because if it is not used properly, it can actually be misused. And we said as a leader, then it's our responsibility that even before the governments and other bodies really come and start bringing in legislations to actually manage that because it is all new. We said we who really understand what AI is all about and what AI can do and what it all can't do. We might actually be in a position to actually develop what is it that we will do and what are the boundaries beyond which we will not go. And fractal became one of the first forms to really develop about an ethical approach. And so much so NASCOM and that's the industry body in India of all the software and all the technology companies in India. They have sort of taken a lot of the learnings and requested fractal to actually collaborate. And so fractal along with some of the other forms has actually worked with NASCOM to develop much more broader national level ethical policies about AI. So AI is very, very powerful. But if it is not managed properly, it can go beyond the guardrails and it can cause problems. And right now as a responsible company, we actually looked at in our cell regulation and that's what our ethical AI policies all about. - It's very interesting. We start looking in some of the companies that we work with in particular a criterion. Also focus on the not only the technology, but the training that's associated with that to help that. Now I help response will use, but also making it possible to correctly use the technology to achieve the result you want. So there's the ethics, but there's also the efficacy that needs to happen. Are you involved in training, upskilling, re-skilling? I imagine with the data analytics and the decision-making that has to be made, the people using these technologies need to be trained differently, need to understand differently. Then what they had done before in the past. - Yes, I think the policies are obviously as made known to every other employee within the organization, but one of the organizations that we recently acquired that is actually is the leader in education and that's for the data analytics and AI is concerned in India. And even that company is really evolving. It's a right from, you know, so this is whole education about AI and ethics as to us is a much wider, you know, belief. And we really lift that belief. Obviously every client of ours that we work with, some of them already have their own ethical policies. So we closely work with them, but we're also helping a lot of our clients to actually set up their own ethical boundaries of AI. So that is actually one of the significant practices of a factor which is helping companies to actually create their AI ethical framework. - And how about in training their employees for, you know, beyond the ethics just to be able to use some of the new technologies? I imagine for some of your clients or potential clients that not only is this technology exciting but it also is puzzling in do they have the workforce that's able to utilize that technology properly? - Yes, I think we, that interaction with our clients have been at different levels. For example, COVID that didn't happen otherwise every year we have meeting where senior most practical employees and senior most members of our clients come together and some of the experts. experts around the world on AI and related fields, the behavioral science and design come together. And that's a two to three day scenario where we actually discussed some of the most cutting edge things. We in Lord, we have client wise specific discussions and interactions wherein we share some of the latest findings. For example, as I said, my team has worked recently on some of the new learnings and decision making from neuroscience, most on smartphones. And we've come up some very interesting new facets around there. We have shared that with our clients so that their digital marketing capabilities can be improved. So we are always on a constant sharing mode with our clients because it was it's a client first. It's a very big philosophy again within within fractal. So whatever we know, we should do tell us what they keep. Now we are constantly looking for new ways to their problems. One of the interesting things that we see on a regular basis these days is that, you know, I don't know if you saw this popular meme that says that, you know, don't fear AI, you know, for AI to take over, people should understand their problems and most people don't understand their problems. I don't know if you saw that. So basically, saying that you can't really do much with AI unless you really understand your problem at its core. How do you feel about that problem solving component with AI and human interaction? You know, where does it begin? Where does AI play a role and how much human understanding needs to be in that process to be able to really solve a problem? Oh, that's an interesting question because it's also because that's clearly an area when the client comes to with a problem. I think one of the things we really look for and says, is that really the problem? And many times we actually use some of our deeper neuroscience and behavioral science principles to go a bit more deeper into the human decision making process to understand that. And then many times, I think we look to sort of reframe that particular problem. So reframing the problem is clearly in a term, not that every time we need to do that, but we do focus to see is the can the problem be redefined and by redefining it, I know ensure we are definitely then trying to solve the problem in a far more efficient way. It's definitely interesting. I think there's perhaps, you know, if you rely too much on the AI, you lose out on some of the creativity and some of the opportunities there. I think many people will feel, especially right now, if you think about students that are in college, you know, they're sort of in this mode where they're just plugging straight into chat GPT and things like this. And winds up happening because they're getting these very cookie cutter responses to things and they're using that to learn. And, you know, I just think about, you know, how much more can they learn if within the classroom, the teachers incorporated chat GPT and help them, guided them to actually using the tool to analyze and understand something versus just sort of regurgitating information. Exactly. And we know not much before AI really came about, there was an industry which is the advertising industry wherein the inputs, which was the, what we call the creative brief that was given to the creative team and that creative brief is prepared by the strategic team. That many time decided what could be what might will be the output that the creative team actually came up with. Similarly, now in a chat GPT there, but it depends on the prompting and the prompting strategy. And we believe there's the real creativity of which at a generative AI is really going to come not from generative AI, but from the prompts that we are going to really have to go to develop. And at fact, we strongly believe that that's again, an area where behavioral science, neuroscience is really coming in to actually help understand the decision making process that happens in the brain and use that to actually develop what we call very creative prompts so that the outputs can actually be far more creative. Otherwise, you give the same prompt, I give the same prompt. The GPT might almost give the same output. Now there's no differentiation, but we know business success is all about differentiation. And how do we create differentiation? And I think we focus a lot on by focusing on by creating a very creative prompting strategy. Yeah, it's almost as if you know, if you're if you're if you just go into chat GPT and say, Hey, how can my company make money? You're just going to get a bunch of suggestions that are not very relevant to you. But the more specific you can get, the more context you can provide it, the more details that you're going to get. And the more exact, exactly, exactly. I think that is where I think a deeper understanding of the problem allows us to bring in all that into the prompts. And one of the tools that we have developed is called the prompt enhancer. And I think that itself an AI tool. So you give a prompt and the prompt enhancer actually makes that prompt even more what I call a vocative. And we have seen that normal prompts, but we give the prompt enhancer the output that finally generate the way I come so as far better when an enhancer has been used. So that's a big that's an area where fractal is focusing on a lot on. One of the areas that I see you looking in this may relate back to some of the other conversations was you've done a lot of work in India on a variety of different topics coming out of the pandemic, COVID-19 work, looking at adoption rates. And then even trying to study the demand for misinformation, which I found particularly interesting in the background of the company. And I know you did it around misinformation around the vaccine and declining vaccine competence and things like that have probably interesting to hear a little bit more about your application of that. But have you thought about taking that broader or doing more with that with I'm sure not only in India, but all around the globe, certainly in the United States. And in Europe, there are lots of battles with misinformation from time to time, whether it's government sponsored or one individual that tends to say something that catches fire and everybody is starting to repeat it. Yeah, I think actually it was a global study to that extent. We, we, it's a multi-country study on vaccine hesitancy. United States was one of them. In India, another Pakistan was another. Then we had four countries in Africa. So, so to that extent, it was a much wider possibly one of the largest studies that have been done on why did people actually not take the vaccine. And that was done by our behavioral science expert company, which we became part of fractal about five years back final mile. And what we actually found was we had some very interesting insights. But one of them was that the whole decision was we thought it's a health related decision. It wasn't many a time it actually became a political decision. For example, we know what happened in the United States and what was the political establishments attitude to was vaccine. And that did affect the people's attitude. Or for example, in Pakistan, there was a change of a coup and there was a change of government. And one of the things we found was that that decision to take a vaccine or not, depended a lot on the trust people had in the political leadership. It had nothing to do with the health workers, nothing to do with health, you know, any other facets of the health system. It had to do with the trust that people had in the political system. And across the world, whichever countries where people's trust in the government was very high, we found the vaccine vaccination levels were much higher. So it's very interesting. That's the point that we set about human decisions. It's a very complex and interesting perspective. And once you take a wider, you start getting this very interesting insights about human decision making. It makes more challenging to train. I, if you're trying to get it to follow human decision process, which can change so quickly. I think that's one of the areas. is I think fractal is working on saying that yes, this vaccine hesitancy project, we didn't really have much of AI out there because yeah, through and through. But one of the things that drives our work is that human brain is actually the most energy efficient machine that the universe has ever seen. And today, you know, and this machine is just working at 20 watts. So if we can really figure out the algorithms that's happening between our years and then take that some of this understanding on to the machines, we might we would end up actually making more energy efficient or overall far more efficient machine. So, so for us neuroscience and AI can actually collaborate with each other. Obviously, a lot of learnings AI can take from neuroscience and we also believe that AI will be able to help neuroscience understand human brain far better. Definitely super interesting and I think, you know, every day we're sort of waking up to new new news about AI and what's going to be happening. My final question for you is, you know, when you think about the skills that are needed for your companies and your clients, to be able to operate at a high level incorporating the human behavior aspect to AI aspect, what are those skills that you see that are relevant for companies or these leaders within these companies to really learn or even the practitioners? See AI is the new field and AI is something which is really clearly we know that's where the future will be. I think one deeper wider quality besides saying engineering and you know mathematical skills or statistical skills, I would say it's an attitude and attitude is I think one willingness to embrace new things. You know, go beyond the status quo. That's one. Two, this is a constantly evolving area and not not only we should adopt, but I think we should be willing to learn on a regular basis. So, lifelong learning is what AI is going to demand of anyone who's going to be involved in that. And the third is I think we should also have a multi-disciplinary approach. We should not be looking at the world of AI only through the lens of AI, but I think we have to look at it from all other sides and that's what we said fractal. We look at it from there are engineers who are looking at the engineering back in the foundation of the AI. Obviously, there is the behavioral science and design which is actually really allowing the interaction of that with the human beings. So, I think we have to look at it from all multiple angles. I think that is what is going to make in any organizations AI strategy really powerful and really effective. Was any final thoughts? Just in following that is for companies to start looking at how they develop their talent. What talent they're looking for is to keep some of these the same thoughts in mind is that AI is happening, it will happening, but it's sort of the following the evolution of that looking for people that are open-minded to not only embrace it, but also that are willing to put in the work, the training, the upskilling, the rescilling, in order to be able to take advantage of it properly. Exactly. I'd everyone appreciate your time, B.J. We'll connect again soon. Take care. Yeah, thank you. Thank you very much. Good to meet you all. Yeah. Thank you. Thank you. Thank you for listening to the Certified Skills and AI and Tech podcast. Brought to you by Criterion, where we provide platforms, software and test development and and delivery services for certification programs.