Hyper-Personalisation to Agentic AI: The Next Phase for Banks | Newgen | The Fintech Show #160

AI isn’t new, but the computing power now available is turning long-promised ideas into practical, bank-ready solutions. In this conversation, Newgen frames today’s moment as the shift from experimentation to impact, with 2024’s pilots becoming 2025’s production deployments. The through-line: start with specific use cases, embed AI into real workflows, and keep humans in the loop so speed doesn’t come at the expense of trust.

The first theme is hyper-personalisation at scale, as what used to be “private-bank” treatment for high-net-worth clients, tailored products, dynamic pricing, proactive advice, is being extended to the mass market. By analysing behaviour, goals and transaction history, banks can recommend more relevant account types and offers rather than forcing customers into generic products. This same thinking underpins broader revenue plays: monitoring interactions to surface the next best product with higher likelihood of adoption, and spotting early signs of delinquency so teams can restructure debt before it becomes a collections problem.

Access to credit is another focus where AI can help as traditional scorecards and bureau data can exclude thin-file borrowers, whereas AI-driven underwriting can incorporate alternative signals, widen eligibility and fine-tune pricing and terms to match a borrower’s real risk. The result, however, isn’t a free-for-all; it’s a more nuanced decision that expands inclusion while managing exposure.

On the efficiency side, intelligent document processing has moved from OCR to “read, reason, and summarise.” so instead of slogging through 400-page reports, underwriters receive concise briefs and draft credit memos in minutes, with humans validating rather than originating from scratch. Similar productivity gains also appear across onboarding (KYC verification and data entry), lending (collecting and validating missing documents), and compliance (tracking regulatory changes, reviewing contracts, monitoring interest-rate caps) and the pattern is consistent: AI does the heavy lifting and people handle judgement.

Newgen argues that how banks adopt AI matters as much as what they adopt; most clients pursue a use-case approach, infusing AI into existing journeys as bolt-ons, delivered through agile, sprint-based programs that control scope, cost and risk. Modernisation is unavoidable though: data silos and monolithic cores raise the price of change, so institutions are deconstructing cores, wrapping legacy with digital layers, and moving toward cloud-native, microservices architectures. Newgen positions its low-code platform and AI studio as accelerants, drag-and-drop components, pre-built models, containerised services, and the ability to deploy selectively at scale which is supported by onshore/nearshore/offshore teams and a large partner ecosystem.

Successful programs share a few traits as they start with a well-defined business problem, integrate AI as a co-pilot with clear handoffs, and bake in governance and explainability from day one. Cultural change is often bigger than the tech shift: teams need transparency to trust the system, and leaders need the discipline to “fail fast” when something doesn’t work. Interestingly, smaller and mid-sized banks often move quicker because they carry less technical debt.

Looking ahead, the conversation contrasts rules-based automation and predictive models with agentic AI that can plan and act across multi-step processes, say, assembling evidence, drafting summaries, and recommending next actions on a loan. Over the next five years, Newgen expect agentic capabilities to be woven into daily banking, with routine tasks automated, products that adapt to customers in real time and more proactive experiences. Cloud adoption, embedded finance and super-app integrations will continue to thrive, while regulators globally sharpen expectations around security and data privacy.