Outcomes over headcount: Genpact's agentic pitch to UK banks

Genpact has rarely been the loudest voice in the UK business process services market. For a business of its scale, exceeding $5 billion in annual revenue in 2025 with over 140,000 employees globally and a financial services book that alone represents over 25% of its global revenues, its brand presence in this region has historically been understated. That is, at least in part, what Moonish Sood, Genpact’s Financial Services lead for Europe and Asia-Pacific, based in the UK, is here to change. Appointed to lead the business last year, Sood brings a deep operations background and strategic clarity that comes across clearly when you sit down with him.

I spent some time with Sood recently discussing Genpact's approach to the banking market and how it is looking to take advantage of the opportunities afforded by the latest wave of tech innovation. Two clear themes dominated the discussion: the maturity and momentum of Genpact's advanced technology capabilities, especially as they build their leadership in Agentic Operations, and the challenge of converting a sometimes-sceptical UK market into a growth engine for the business.

Genpact was a central player in creating the business process outsourcing (BPO) category back in the mid to late 1990s, initially as a General Electric 'captive'. It has now, in Sood's telling, retired BPO. The replacement construct is ‘Agentic Operations’ - a shift from human-led, human-validated processes to machine-led, human-validated ones. The shift is simple to understand but will have huge consequences for operations: where previously humans executed and humans validated, the new model has machines executing and humans validating. The ultimate direction of travel, Sood is clear, is towards humans handling complex exceptions rather than the processing side of things. Handling the complex exceptions is Genpact's sweet spot given its domain and process intelligence muscle.

Many of Genpact's clients use its agentic products globally, with finance and accounting (F&A) including accounts payable, procurement, insurance, and banking as the most mature solution areas. In F&A, the numbers Sood cites sound impressive with invoice processing costs having halved. In insurance, Genpact’s Insurance Policy Suite targets the commercial underwriting workflow from submission to bind, with projected outcomes of up to a 75% reduction in cycle time, up to 50% lower operating costs, and up to 30% more underwriter capacity. In banking, Genpact’s Banking Analyst Suite shows a similar profile for in-scope anti-money laundering (AML) alert investigations: up to 80% lower handling time and up to 40% lower total cost of ownership, with a documented rationale and audit trail behind every recommendation.

Indeed, the latter is a good illustrative example of where AI-driven triage directs human attention to the highest-propensity fraud cases, rather than requiring analysts to work through a queue sequentially.

The financial services sector Sood describes is more nuanced than the headline opportunity might suggest. The assumption that banks, as technology-confident and well-capitalised institutions, would be early movers on agentic adoption turns out to be only partially correct. Front-office and consumer-facing technology has seen enormous investment and banking apps are, as Sood notes, effectively technology products now. The back office is a different story. Legacy infrastructure, accumulated through decades of acquisition, remains a drag. AI governance structures are immature. Many institutions are still cycling through individual proofs of concept without having established the common AI architecture that would allow them to scale.

Seeing active demand for agentic AI services globally from Genpact's vantage point, is - Sood references a Banking Analyst Suite implementation with Banks in Australia, another with a newly signed deal bringing both F&A and Banking Analyst Suite into an agentic model of delivery. Europe is moving quickly, particularly in financial crime. Genpact has recently completed what Sood describes as a significant rotation of an existing European bank's financial crime operations into an agentic model, with a major European payment services provider following a similar path.

The UK lags. There are conversations, and its relationship with a ‘tier one’ bank, where Genpact built an AI centre of excellence, helped prioritise the project pipeline and is now delivering, represents an important reference point. But Sood is candid that UK banks are generally slower to commit. Regulatory caution and the Discretionary Commission Arrangement (DCA) remediation uncertainty around auto finance are suggested factors. The agentic shift is also, he notes, a very different kind of change management challenge from prior technology transitions, partly because it lands directly in the operational core of an organisation and requires internal champions at the client to drive it forward.

Genpact's go-to-market approach is also evolving under Sood's stewardship. The business has historically been focused on mining its existing client base — predominantly large global enterprises. Sood has introduced a more systematic focus on opening new large enterprise relationships in markets where Genpact is currently absent (other tier-one banks are mentioned by way of example). A third approach targets what the firm calls 'next-gen enterprises', institutions below $10bn in revenue, including fintechs, building societies and specialist lenders.

The commercial model is changing too. Genpact is explicitly moving away from full-time equivalent (FTE)-based contracts towards outcome-linked commercial structures, with an aspiration to reach 100% non-FTE-based revenues quickly. Genpact’s proposition is underpinned by delivering outcomes such as reduction in cycle time, increase in market share, broker satisfaction, or improvement in customer net promoter score (NPS). Sood acknowledges that clients are drawn to the narrative but that translating excitement into real change means change in the clients’ operations, and that’s a harder thing to achieve.