CNME Editor Mark Forker secured an exclusive interview with Tomas Skoumal, CEO and Co-Founder of Dyna.Ai to learn more about how the Singaporean-based AI-as-a-service company plans to transform customer experience in the financial services sector – and how their portfolio of solutions are empowering human operators to deliver better levels of CX through the power of LLMs and machine learning.
Tomas Skoumal is a banking veteran with over 20 years of experience at some of the biggest names in the world of corporate finance.
Citibank, RBS, and Standard Chartered are just a few of the financial behemoths that Skoumal has worked for in what has been an illustrious and distinguished career to date.
Skoumal has made a successful career out of helping his customers manage and assess risk.
However, in 2024, Skoumal decided it was time for him to take a risk.
Despite the relative comfort, security, and financial trappings of life working as the Global Head of Loans at Standard Chartered, it was time for a new challenge.
Skoumal decided it was time to enter the world of entrepreneurship.
In April 2024, he formally established Dyna.Ai, an AI-as-a-Service company that is headquartered in Singapore.
The mission statement of the company is to transform the finance industry globally through the adoption of cutting-edge AI products and solutions.
CNME sat down with Skoumal to get a better understanding of the business model of Dyna.Ai, and the primary factors that led to its inception.
One of the areas that Skoumal is looking to transform is an area that he knows all too well from his banking days, and that is call centres.
“I have spent 20 years working in financial and banking institutions, and most of the time I was tasked with the responsibility of managing risk and sales, and in both of those respective positions a big part of the job remit was focused on managing call centres. When you’re managing those call centres you do face a plethora of issues, such as the turnover, growth, space, hiring, products and so on and so forth, but it was a component within the overall business unit where you always trouble shooting and trying to overcome issues to deliver high-quality CX, and it was an area of the business that I enjoyed,” said Skoumal.
In 2017, Skoumal was working for Chinese consumer finance company Home Credit.
During that role is where the seeds for his AI-driven Contact Centre solution portfolio at Dyna.Ai were sown.
“In China, we had 12,000 collectors, but I was working directly with a local Chinese company on pre-recorded answers that was based on content understanding. We were able to use this content understanding to decrease the volume of hours in relation to 12,000 collectors within the timespan of 3 months down to 4,000 collections – and that for me was a powerful demonstration of how we can use technology to reduce costs within call centres and make them more efficient and effective. For the last 5 years of my corporate career, I was hyper-focused on reducing overheads with call centres and ultimately making them more standardised. Call centres was an area that I always enjoyed working in despite the challenges we regularly faced. In addition to this, I was acutely aware of the potential that was there within this space because I had been so close to that side of the business in my previous roles,” said Skoumal.
In terms of what triggered Skoumal to take the plunge and go into business on his own, he was candid in his revelation that he simply wasn’t suited to his role at Standard Chartered, and that ultimately the bureaucratic nature of the job led him to scratch the itch of owning his own business.
“Look, I had forged a good career for myself in the corporate world, and I was well-established and respected by my industry peers, but I did always have this burning ambition to run my own company. However, when you have a wife and two kids it’s not easy to get out of the rat race as they say. I was headhunted by a few persuasive recruiters, and they convinced me that it was time to go back to the corporate side of banking, and that’s what I did. I joined the Global Office of Standard Chartered, but I didn’t like the bureaucratic nature of the role to be honest. It just didn’t suit me, given my background in sales I was always a man of action. I was accustomed to having my own portfolio, where I either wanted to make more money on the portfolio, or simply try to manage the portfolio better, or whatever the KPI might be, but I was driven by tangible outcomes, and that inspired and motivated me. Whereas at Standard Chartered the job remit was just not suited to my personality, and that experience ultimately led me to go out on my own, and start Dyna.Ai,” said Skoumal.
The conversation then pivoted towards what makes Dyna.Ai different from other market players.
Skoumal said that still being in that startup phase fosters an environment in which innovation can flourish.
“The business that we operate in is a saturated marketplace, there is no question about that, and it is an ultra-competitive landscape. I think one of the elements that differentiate us is the fact that we are still a startup, and that culture of innovation is palpable throughout the entire organisation, and being in the that startup state inevitably allows us to be agile,” said Skoumal.
However, Skoumal was unequivocal about what he believed was the key differentiator in the armoury of Dyna.Ai.
“I think the fact that we have our own LLM is huge. We can implement our LLM and deploy it on the private cloud, but that’s an easy to thing to just say and throw out there, so let me provide you with more of a technical perspective and background. We are using conversational generative AI, and we are using the voice bot to communicate to the end customers of our financial solution customers. We are also helping our customers by leveraging the power of machine learning to help companies tackle the issue of discriminative AI, or biases in AI. We can do scorecards for our customers where we can rate them on their AI biases, then they can use that data and label it, and then the data that you get from every single call in the call centre can be used to enrich that scorecard. That’s very powerful,” said Skoumal.
Interestingly, Skoumal believes that the market monopoly currently cornered by US chip giant NVIDIA will come under pressure from Chinese ICT leader Huawei over the next few years.
He believes more players will enter the GPU market, which will in turn see GPUs become more affordable.
“I do believe that GPUs will become less expensive, and I think the entire process will become cheaper. Other players such as Huawei will enter the market, which is needed, because at the minute there is a bit of a monopoly with the likes of NVIDIA. I think Huawei can lead in this area, and I know that a lot of people will not believe that, but look if you said BYD will produce more cars than Tesla then you would’ve been laughed at, but that is the case – and I think it’ll be the same when it comes to GPUs between Huawei and NVIDIA,” said Skoumal.
As aforementioned above, Dyna.Ai deliver AI-as-a-service, and are hyper-focused on transforming the contact centre space through the deployment of AI solutions and products.
Skoumal said that when it comes to AI in a CX setting there still needs to be a mindset shift.
He highlighted that we are much more forgiving of a human error than that of a mistake made by an AI agent or automated chatbot.
“Many of my friends that use voice and chatbots when interacting with certain organisations find the process in many cases incredibly frustrating. They have become so accustomed to the traditional way of doing things, and they don’t fully understand or appreciate the real purpose of what the types of questions that the agent is asking. We are using LLMs when it comes to the agents that we deploy, and we are basing it on the similar training that human operators would go through when being onboarded as a call centre agent. As humans, we have much more tolerance to human mistakes than those made by machines. If you are talking to a human operator and you’re put on hold and they are making mistakes then research shows that we are much more willing to forgive, whereas in contrast if we have the same experience with an AI agent that we are far less forgiving,” said Skoumal.
Skoumal concluded a superb discussion by highlighting the two-fold role AI plays in thwarting online fraud and theft.
“I think when it comes to using AI to combat fraud it’s not generative AI, in fact, it is good old fashioned machine learning. However, the generative AI component can drive the machine learning scorecards and deploy them accordingly. When you are going through the process of attempting to detect fraud you typically go through some form of machine learning. However, going through these processes on a one-by-one basis is extremely difficult, but this is where the AI can come into play. If there some sort of fraud hitting the bank and there are 10,000 transactions then you human operators can’t call 10,000 customers immediately, but the AI agents can do this. AI agents can also ask the relevant questions and better navigate the customers as to what to do next in terms of blocking that transaction or explaining how the transaction was done. So, in summary the AI can play two key roles, the first of one is preventing the fraud, and secondly when it happens, they can contact the customer immediately and give them the feedback and support they need to resolve the situation, and that’s what we do at Dyna.Ai,” said Skoumal.
