Building a distributed agile/digital first model

Successful programs in financial services deliver digitalization or automation by using an overarching approach, not just a bundle of projects tracked for their progress in red, amber and green. The following summary refects on learnings from several approaches and rests on the assumption of a built-in superiority of a “digital first” approach. As a north star, any activity should be as digital as possible within a short period of time (18-36 months, as an ambitious start).

If taken in pure form, this approach leads to the convergence of operations and technology in 4 phases:

1.??? Shift and drop of services: Operations Processes and IT Services are shifted from a location to a “Operations and Technology hub” as they stand. The location of the hub can be driven by considerations of talent pools, strategic partnerships and/or factor costs. As an explicit choice, no fixing of a service at origin is envisaged. It is crucial to define service levels, if not done so, and “freeze” a baseline in the point of shifting in order to be able to measure improvements driven by the model.

2.??? Optimization along production models: In a second step, the services which are aggregated into hubs are optimized to deliver economies of scale and scope. The focus lies on simplification. The business recipient now has to describe expected improvements, e.g., service levels and consumption efficiencies by concrete activities, and translate these into decision trees.

3.??? Convergence of Operations and Technology: The third step works both as “shift, fix and drop” and “Fix, shift and drop” as long as the first and second steps are in swing. The objective is to deliver impact with rapid pay-back periods and a tangible return-on-investment within 12 months, but involves more than simple personnel cost reduction. As a start, inbound customer or client requests (e.g., AML/KYC) serve well as a pilot. “Automate to generate” – before the last step is initiated, it is critical to lay the ground work by automation first (“learn to walk, then run”).

4.??? Ambition Generative AI/digital first: With regard to the target state of Operations Processes and activities, the objective is to push ahead with Generative AI. This will require a re-design of processes radically following a “digital first” philosophy. The hubs are the nexus of ultimate or “real” convergence of Operations and Technology, i.e., replacing activities by Technical Services. Given the technological developments of the last 24 months, pre-trained models will be sufficient to cover level one (basic requests) and level two (standardized requests tailored to profiles) decisions.

Any requests beyond level three, such as individual engagement and unique experiences for the client, are of greater concern but will be solvable by creating a small own model heavily leveraging available commercial models and training on specific company data in a secured space. Building a quality-assured specialized data set is here central piece of the strategy and pre-requisite for regulatory approval, and for client experience.

***

Beyond the direct impact of value creation from labor arbitrage, process optimization/simplification, automation and “digital first”/Generative AI, the model described here seeks to also open the gates to value drivers of the future. As a built-in feature, the hubs enable innovation, especially around data, by creating network effects and out-of-the-box-thinking (e.g., from fund accounting to data analytics). The resulting access to capabilities and idea metabolism enables the organization to digest future trends into quick proofs-of-concepts.