
No Free Lunch With Greg Stewart · 2026-03-06 · 15 min
Key moments - from our scoring
Substance score
57 / 100
Five dimensions, 20 points each
Old Mutual's approach to digital transformation prioritizes making financial services genuinely accessible to underserved populations in South Africa. Ramsamy highlights SmartCoach, a multi-agent AI wellness assistant within the MyMutual app that engages customers within their actual portfolio context, helping them arrive at financial advisor conversations better informed. The platform has driven over two million financial services activities completed in the past year. Beyond consumer tools, Old Mutual Bank targets South Africans earning R1,000 - R30,000 monthly through completely digital platforms, while the insurance division deploys AI for high-volume servicing and underwriting with a methodical, lower-risk-first approach. Ramsamy stresses that responsible deployment matters: Old Mutual navigates POPIA regulations, implements guardrails to prevent bias in large language models, and uses voice-to-insight AI solutions in Kenyan call centers for real-time training. He argues that 80% of large financial services institutions globally will have adopted generative AI by year-end, yet many remain stuck between experimentation and execution. Success requires not just technology but strong leadership, cultural alignment, addressing workforce fears about job displacement, avoiding innovation theater, bridging the middle management bottleneck, and building genuine data culture - not just infrastructure.
SmartCoach is Old Mutual's multi-agent AI wellness assistant within the MyMutual app that engages customers in the context of their actual portfolio, financial goals, and personal circumstances. It uses guardrails to prevent bias and recommend actions, helping customers arrive at conversations with financial advisors better informed and more confident.
Old Mutual operates within South Africa's POPIA regulations, which were strengthened to cover breach reporting, consent for direct marketing, and cross-border data transfers. The company implements guardrails to prevent bias in large language models, conducts checks to prevent irresponsible use of personal information, and ensures service providers meet responsible AI standards.
Old Mutual Bank targets South Africans earning between R1,000 and R30,000 monthly - a segment historically underserved by traditional banks. It uses completely digital platforms to deliver cost-effectively and responsibly, leveraging existing customer data from other Old Mutual products to improve service while maintaining data integrity.
Predictions suggest that by the end of the year, more than 80% of large financial services institutions will have adopted generative AI, up from between 5-10% just about a year ago. This rapid adoption rate reflects intensifying competition and demonstrates that digital transformation is no longer optional but essential for survival.
Key barriers include workforce fear about job displacement, leadership ambiguity and innovation theater (championing AI publicly without embedding it in real decision-making), middle management bottlenecks, trust deficits around AI-driven decisions, and gaps between data infrastructure and genuine data culture.
Our reviewer’s read on each dimension, with quotes from the episode.
The episode contains substantive ideas about AI adoption, financial inclusion, and responsible AI deployment, but is hampered by significant transcription degradation that obscures roughly 20-30% of the content with garbled text and incomplete sentences. Where legible, Ramsamy offers concrete frameworks (upstream engagement vs. reactive, cultural vs. technical barriers to AI adoption, data culture vs. infrastructure), but the execution feels somewhat high-level and repetitive rather than densely packed with novel claims.
It's this shift from kind of reacting to problems after the fact to kind of anticipating them instead of engaging with the customer a claim or complaint or after a lapse after a financial shock, you actually using models to engage upstream
It requires more than just good technology. It requires like strong leadership, the culture that goes along with the adoption of like a pervasive technology which is AI.
Ramsamy presents some counterintuitive thinking - especially the emphasis on cultural and leadership barriers to AI adoption over pure technology, and the distinction between data infrastructure and data culture. However, the core arguments (AI is essential, responsible deployment matters, inclusion through digital tools) are familiar industry talking points. The SmartCoach example is concrete but presented more as a case study than a fresh insight.
It's not cloud computing where it took us quite a while to be able to show business what cloud computing, the value of cloud computing and things like that. AI is more general tool as well.
How do we make sure that we're not furthering bias that might be built into those models?
Ramsamy is credible - he holds a Group Chief Technology and Data Officer role at Old Mutual, a major African financial services conglomerate, and has clearly shipped real products (SmartCoach, AI claims handling, call center analytics in Kenya). He speaks from operational experience at scale rather than theory. However, the transcript does not establish depth of hands-on execution or provide timeline/scale details that would elevate him to top-tier caliber.
Desen is the group chief technology and data officer for Mutual Limited
we launched inside MyoMutual... It's like it's the country's first like multi-agent wellness assistant
The episode suffers from severe lack of specificity despite a few concrete data points. Named examples exist (SmartCoach, Aviva's 23-day liability assessment reduction, Kenya call center solution, two million financial services activities), but most claims lack numbers: no customer satisfaction metrics are actually cited, retention improvements are asserted but unmeasured, and the South Africa digital bank targeting 'a thousand to thirty thousand' earnings has garbled context. The transcription quality makes it difficult to assess whether specificity was present but corrupted.
Aviva in the UK... they've managed to cut liability assessment for quite complex cases by 23 days using predictive modeling and AI
over two million financial services activities completed in the past year
Stewart's questions are straightforward and topical but rarely push back or challenge. He asks open-ended setup questions ("Why has digital transformation become essential?", "How is this transforming client experiences?") but does not follow up on vague claims, ask for numbers when metrics are mentioned without values, or probe tensions (e.g., how to balance inclusion with data minimization). The interview reads as a friendly walkthrough rather than a substantive interrogation. No productive disagreement or sharp follow-ups.
From your seat group chief technology and data offset on mutual, why has digital transformation become not just important, but absolutely essential for survival and sustainable growth in today's financial world?
And what key metrics such as rates or satisfaction level scores show benefits ⁓ segments?
Computed from the transcript - who did the talking, and the words that came up most.
In the latest episode of “No Free Lunch,” host Greg Stewart engages with Dhesen Ramsamy, the Group Chief Technology and Data Officer for Old Mutual Limited, to explore the profound impact of digital transformation in the financial services sector, particularly in South Africa. The discussion centers on how digital transformation, driven by AI and rising customer expectations, is reshaping the industry landscape.
Transcribed and scored by The B2B Podcast Index.
Greg Stewart: Welcome to another edition of No Free Lunch, Africa's freshest business and tech podcast with me, your host, Greg Stewart. Our topic today is digital transformation and particularly what the impact of this process has been in the insurance and finance sector in South Africa. Joining me today is Desen Ramsamy. Desen is the group chief technology and data officer for Mutual Limited.
Welcome to No Free Lunch, Desen. Dhesen Ramsamy: Thanks very much, Greg. It's good to be here. Greg Stewart: So Desen, we're living through one of the most disruptive periods in the financial services industry.
And a lot of that has been driven by digital transformation and AI acceleration, rising customer expectations for seamless digital experiences, and intense competition from both traditional players and Nimble FunTechs. From your seat group chief. technology and data offset on mutual. Why has digital transformation become not just important, but absolutely essential for survival and sustainable growth in today's financial world?
Dhesen Ramsamy: Yeah, so ⁓ I it's on a few fronts and the first has to be the use technology to drive better outcomes for the people we serve, for our clients, for our advisors, know, for the markets we play in. So at the heart of it, it has to come down to better service provision, better products. ⁓ know, better experience for these markets. It not only helps you be competitive in those spaces, but should help you those moments of, ⁓ in those moments of to not let down, client or an advisor or whoever you're doing business with.
⁓ And for themselves, ⁓ technology has fundamentally helped to drive down the interaction ⁓ to reach markets that you otherwise would not have through more physical and traditional means. So it is not, we no longer, or we should no longer be having conversations about if, but rather how and how we reorganize and actually deliver digital transformation. Greg Stewart: That's quite important about actually delivering. And in terms of that, how is old mutuals push into AI and data-driven tools like personalized financial platforms?
has that transformed client experiences in South Africa? And what key metrics such as rates or satisfaction level scores show benefits ⁓ segments? Dhesen Ramsamy: I think ⁓ our point, again, ⁓ very simple, ⁓ it down to what we just spoke about It's how make financial services genuinely accessible. How do we please at key moments of truth, it's a claim, ⁓ whether it is providing a all of those things, ⁓ or service issue.
accessibility something like, you know, ⁓ don't shy from because historically a lot of people have been outside, that kind of inclusion of quality financial advice. use that example. So South Africans have a ⁓ knowledge gap when it comes to wellness, for instance. People want to make good decisions about money.
But the tools, the language, the access points. they've been designed for very narrow segments of the population. So a key metric has to be the extent to which we've included people that we haven't necessarily done before. why our AI push is really about closing that gap.
And a good I wanna share here is smart coach, ⁓ which we launched inside MyoMutual. It is ⁓ the MyoMutual app that is. It's like it's the country's first like multi-agent wellness assistant. And what makes it like very different from generic chat pod is it engages customers in the context of their actual portfolio, ⁓ their their goals, their financial goals that is.
⁓ So have agents ⁓ that intent. have agents ⁓ enforcing like guardrails and devices, one for ⁓ recommending actions, for screening And it helps people who use it show up for a conversation with a financial advisor a lot better informed and more confident. we're building this in what you accurately pointed out is kind of one of the fastest moving periods in ⁓ history. And legislation and regulation is catching up to the of the technology set.
So we had figure out how to ⁓ deploy this responsibly. Greg Stewart: Yes. Dhesen Ramsamy: take it through a very set of guardrails in the business. On the insurance side, for example, the All Mutual team is ⁓ advancing a GEN.
AI focused on high-volume servicing interactions and underwriting use cases. And they're taking a approach. They start with ⁓ lower risk, basically process but building out towards more complex applications. you know, as confidence and the capability grows, we get more police clients, we get shorter, claims handling times.
yeah, customer metrics like, you know, NPS, customer engagement scores, ⁓ all of those things are what track when we look at how we drive that out. ⁓ Then beyond South Africa, in our region's business, AI in the customer experience space is at early stage, but there's some practical deployments worth mentioning. In Kenya, for instance, we have a solution running in our call centers. It converts voice actionable insights in near real time.
And it's used for call center training and for surfacing customer experience issues a lot quicker than traditional analytics would have ⁓ So like on the measurement side, what I mentioned is customer behavior with SmartCoach, we can how customers interact with their portfolios, whether they're complete. financial wellness activities, whether they're following through on the goals they're setting. it's tightly linked with all mutual rewards, which has seen over two million financial services activities completed in the past year, which tells you something about appetite, right?
The moment you ⁓ use technology to deliver something that is far easier to use, it drives better outcomes. And our advisors are giving us feedback that, you know, it's the quality of the conversations that are fundamentally changing. Greg Stewart: Absolutely. Dhesen Ramsamy: So better retention, better outcomes, better trust.
Greg Stewart: Yeah. Yeah, I think generally system that gets simplified and becomes often I am talking from personal experience from things like banking and having a headache using an is the worst possible experience ever. It just becomes a no-no. So I think from that point of view, seeing that ⁓ increase or improvement Dhesen Ramsamy: Yeah.
Greg Stewart: a positive thing. Just in terms of data privacy, there's a lot of questions out of data privacy, not just in South Africa, globally, particularly with AI and the ability of AI ⁓ dive into details ⁓ maintain personal details. You to talk to us a bit about how digital transformation is bridging gaps for clients and how ⁓ you're dealing with issue of privacy of data ⁓ where ⁓ Old has tech innovations to improve inclusion, but also to improve data security. Dhesen Ramsamy: so ⁓ Africa has a very specific operating context around that.
Europe does as well. You can't walk away from these things when trying to drive digital transformation. It cannot be, you know, ⁓ irresponsibly. we have Papua, is which was again, I think, strengthened last year, the regulation ⁓ around breach reporting, consent for like direct marketing.
for cross border data transfers. So the regulator is certainly keeping up there. And ⁓ a good thing. It helps ⁓ like ours raise the bar for how we handle personal information because a promise give to ⁓ and advisors.
the same time, you have a population where kind of, you know, digital access is uneven. ⁓ Sometimes is not ⁓ A large portion of the populace that we serve have either been underserved by traditional financial services or sometimes completely excluded. So you can't just build digital products and assume it's going to reach these people. And when you do build it, you need to ensure that you're serving them as responsibly ⁓ possible.
for ⁓ in space, Mutual Bank, approaches to solving been ⁓ completely digital. you know, we're targeting South Africans earning between like a thousand ⁓ thirty thousand a And it's a segment that's been historically kind of underserved by traditional banks. Like our platforms just to deliver cost effectively, but to work responsibly with the information we've already garnered in this space from our interactions with ⁓ on the other side of all mutual. So you have a funeral policy.
How do we use that information responsibly? How does that information responsibly feed into modeling? How do we make sure that we're holding the line our service providers, whether they are giving ⁓ the large language models? How do make sure we're not furthering bias that might be built into those models?
⁓ in series of checks ⁓ to ensure that we prevent the of personal information irresponsibly. ⁓ And the second, is that we also carrying this responsibility to ensure the models we use don't have a baked in and things. Greg Stewart: just in terms of looking a bit as digital transformation is likely to accelerate even faster as we see AI developments, genetic developments also accelerating. us how you see that landscape changing.
Where is that going to benefit clients? Is it going to make things even faster? Is it going to bring down costs further? How do you see that landscape evolving into the future?
Dhesen Ramsamy: I think my honest answer here is globally seeing the industry at a kind of inflection point. There's a of predictions that by the end this year, more than 80 % of large financial services institutions would have adopted GEN.AI. ⁓ that's up between and 10 % ⁓ just about a year two ago.
So that pace of adoption tells you something about how competitive the landscape. is becoming, especially with the adoption of these tools. I think what's most relevant the markets or mutual plays in Africa in the main and in South Africa more specifically, it's this shift from kind of reacting to problems after the fact to kind of anticipating them instead of engaging with the customer a claim or complaint or after a lapse after a financial shock, you actually using models to engage upstream, ⁓ data to flag risks and guide better before things go wrong.
And that fundamentally changes what it means to be a financial services provider. On claims, there's already a lot of evidence, Aviva in the UK, I always quote this case study, but they've managed to cut liability assessment for quite complex cases by 23 days using predictive... modeling and AI, customer complaints have reduced by orders of magnitude through the use of AI. And that's the journey we are on as well.
The reality is lots of the industry is still stuck between experimentation and execution, which is the thing I told you about earlier on, financial wellness AI solution. That's... telling you that it's gone is the glory. But the full story is a lot of perseverance, a lot of testing, a lot of experimentation over the course of ⁓ and a half, three years.
That team, I'm proud of the team that landed that. So organizations can only close that gap by responsibly experimenting, ⁓ scaling the experiments chosen. there's some inherent risk. So you've got to double down on those bets you made that are showing kind of good results.
It requires more than just good technology. It requires like strong leadership, the culture that goes along with the adoption of like a pervasive technology which is AI. ⁓ AI is not just the domain of technologists now. It's not cloud computing where it took us quite a while to be able to show business what cloud computing, the value of cloud computing and things like that.
Greg Stewart: Yes. Dhesen Ramsamy: AI is more general tool as well. So you get a push from the business too. ⁓ So leadership, ⁓ all of matters.
Greg Stewart: And then a last question for you. I know you've got time constraints, but you're going to be speaking at the Digital Transformation Summit next week in Johannesburg. I think it's on the 11th of March. Tell us what is the single most important message, insight you'll be leaving with fellow C level leaders and digital practitioners?
Dhesen Ramsamy: I think I've been very deliberate about what I want to talk about. And it's the cultural and leadership challenges to AI adoption ⁓ in corporates. And I think it comes down to a few things. It's how we address the fear and resistance at a workforce level.
⁓ to separate signal from noise it comes to things like, is AI going to suddenly you know, take my job, that kind of thing. you know, addressing that sort of fear and resistance from the workforce, then there's leadership ambiguity, you know, and there's this thing about innovation theater, lots of executive teams publicly champion AI, but, you know, how do we embed it in actual decision-making and resource allocation, that sort of thing. know, famously the middle management bottleneck with the adoption of ⁓ technology set like will play up.
Your data culture versus the data infrastructure. A lot of people talk about, ⁓ but we need good, clean data. We need the data platform. But what about your data culture?
How ⁓ you conduct business to ensure that there's continual integrity in your data sets? Sometimes there's a trust deficit around ⁓ AI decisions in leadership. There's talent So I think ⁓ O-Mutual has very practical about drawing all of these together under a leadership and culture theme. And that's kind of what I want to more thinking about.
Not that, you know, ⁓ going there to share all the answers that I have. It's just to present that this is far more than a technology challenge. We should be thinking a lot more holistically. Greg Stewart: Yes.
Decent RAM, Sammy, I think we're going to leave it there. Thank you for your insights. I think it's great insights and I think the whole leadership of AI and these developments into digital transformation are critical going ahead. And it will take a very special kind of leadership approach to actually make it meaningful in the marketplace.
Thank you for joining us and I appreciate the insights. Dhesen Ramsamy: Thanks Greg, it was a pleasure chatting to you.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.