Artificial intelligence (AI) has rapidly become part of everyday business life, but for wealth managers the next phase may prove far more significant than the first. That was the central theme of a recent WealthTech Talks webinar hosted by The Wealth Mosaic in partnership with Profile Software, which explored how agentic AI could reshape operating models across the UK wealth management sector.
Moderated by Stephen Wall, Founder of The Wealth Mosaic, the discussion brought together:
- Stefanos Athanasiadis, Managing Director Investment Solutions at Profile Software;
- Jaidip Banerjee, Head of AI, Data & Architecture Services at Leading Point; and
- Reeves Knyght, Founder & CEO of Knyght Digital; formerly Global Head of Digital Assets & AI Strategy at Coleman Wealth Management.
From assistive AI to agentic AI
The discussion opened by distinguishing between what the panel described as two distinct waves of AI adoption.
Athanasiadis said the first wave had consisted of assistive AI: tools that support employees with tasks such as document processing, report drafting, and information retrieval, while leaving humans responsible for completing the overall process.
The second wave, agentic AI, goes considerably further. Here, AI plans, reasons, and executes entire workflows, with humans supervising outcomes rather than carrying out each step themselves.
“The person still completes the process” in assistive AI, Athanasiadis explained. By contrast, with agentic AI, “the human is there, but it supervises... by exception, not by default.”
The distinction is significant, he argued. Although assistive AI may save minutes within individual tasks, agentic AI has the potential to remove substantial amounts of operational work altogether.
Banerjee agreed that assistive AI is already becoming established across wealth management, particularly in areas such as document summarisation and transcription. But he suggested the next phase will depend less on technological capability than on firms demonstrating measurable business value. “The focus should be on the value delivery rather than focusing on technology itself,” he said.
Strategy before technology
Throughout the discussion, the panel repeatedly stressed that successful AI adoption requires strategic planning rather than isolated technology deployments.
Knyght observed that many firms currently feel pressure to “do something with AI”, leading them to adopt individual productivity tools without necessarily having an overarching strategy. Although these solutions can improve administrative efficiency or support advisers, he suggested they represent only the beginning of AI adoption.
Implementing agentic AI, by contrast, requires organisations to rethink how work itself is organised. Rather than simply introducing new software, firms must determine which activities should be undertaken autonomously, how AI agents will be managed and monitored, and how those agents integrate into existing business operations.
"It needs to be a strategy initially from the top down," Knyght said.
Athanasiadis echoed that view, arguing that organisations should treat agentic AI as a business transformation programme rather than an IT project. This includes identifying priority processes, defining measurable success metrics, and assessing the impact on operational performance and profitability.
Reimagining the operating model
The panellists argued that agentic AI ultimately requires firms to rethink their operating models rather than simply improve existing workflows.
Banerjee described this as "reimagining the entire operating model", noting that many wealth managers continue to operate multiple legacy platforms and fragmented technology estates. Integrating AI into these environments presents additional complexity – particularly where organisations seek to scale agentic capabilities across multiple systems.
Knyght contrasted this challenge with building an organisation from scratch. While an established firm must integrate AI into existing structures, a new business can be designed as AI-native from the outset.
He compared the difference to hybrid and electric vehicles. Existing firms implementing agentic AI may resemble hybrid cars, combining traditional and AI-driven approaches with the complexity of both. New entrants, however, enjoy the opportunity to design entirely AI-native operating environments without those legacy constraints.
Delivering measurable outcomes
To illustrate how agentic AI can be applied in practice, Athanasiadis shared a Profile Software case study focused on fund administration reconciliation.
Rather than attempting to automate simple rules-based tasks, the project targeted a complex daily reconciliation process that traditionally required one to two hours of specialist staff time each day, per fund client. The resulting architecture combined conventional automation, AI reasoning, and deterministic validation – while retaining human oversight for approvals and exceptions.
According to Athanasiadis, the implementation reduced processing time by at least tenfold while delivering cost savings exceeding 75 percent. The project also improved consistency and auditability by capturing operational knowledge within the system rather than relying solely on individual expertise.
More broadly, he argued that successful implementations combine three essential elements: agentic AI technology, access to core operational systems and data, and deep knowledge of business processes. Missing any one of these, he suggested, significantly reduces the likelihood of success. The results were validated to an audit standard across 85 real reconciliation events over five trading days, with every action logged and replayable for audit.
Athanasiadis noted that this convergence is explored in depth in Agentic Triada: The New Economics of Wealth Management, published by Profile Software with IBS Intelligence, describing these three elements as the foundations of agentic success: technology alone is insufficient without the right systems, data and business expertise.
Data, governance, and trust
Although AI capabilities continue to develop rapidly, the panel agreed that technology alone is insufficient.
Banerjee highlighted data as a critical foundation for successful deployment. Before organisations can introduce agentic AI at scale, they need a clear understanding of their data, governance arrangements, security classifications and access controls.
Equally important, he argued, is explainability. Organisations must understand how AI reaches decisions so that human supervisors can confidently review and approve outcomes where required.
Trust was another recurring theme. Although younger generations may become increasingly comfortable interacting with AI, Banerjee suggested many clients still expect to trust an individual adviser.
Building confidence in agentic AI therefore depends on demonstrating reliable, transparent outcomes over time – rather than expecting immediate acceptance.
Looking ahead
Looking to the future, the panellists offered differing but complementary perspectives on where firms should focus.
Athanasiadis encouraged organisations to “avoid the hype” and instead establish structured transformation programmes with clearly defined objectives and measurable outcomes.
Banerjee pointed to data management as the essential starting point, while Knyght suggested wealth managers could initially use agentic AI to reach younger and emerging client segments that are currently underserved – allowing traditional advisers to continue focusing on more complex, higher-value relationships.
The key takeaway: agentic AI represents more than the next generation of productivity tools. Realising its potential will require wealth managers to rethink processes, operating models, and organisational structures alongside the technology itself.
Want to watch the webinar in full?
The full webinar, The agentic manager - How agentic AI is reshaping work in UK wealth management, is now available to view here.
About Profile Software
Founded in 1990, Profile Software is a specialised financial software solutions provider with offices in key financial centers and a presence in 50+ countries across Europe, the Middle East, Asia, Africa and the Americas delivering market-proven solutions to the Investment Management and Banking industries.
Profile Software is recognised as an established and trusted partner by international industry-specific advisory firms.
About The Wealth Mosaic
The Wealth Mosaic is a UK-headquartered online solution provider directory and knowledge resource, focused specifically on the wealth management industry.
For wealth managers, the buy side of our marketplace, The Wealth Mosaic is designed to enable discovery of key solutions, solution providers and knowledge resources by specific business needs.
For solution providers and vendors, the sell side of our marketplace, The Wealth Mosaic exists to support the positioning, exposure and business development needs of these firms in a more complex and demanding market.
