3 minute read 3 Jul 2019
Close-Up Of Gloved Hand Holding Glass Ball

How to optimize your intelligent automation build

By

EY Global

Multidisciplinary professional services organization

3 minute read 3 Jul 2019
Related topics AI Automation Technology

Show resources

Completing an assessment and design phase is crucial to ensure a successful build of an automation program.

Intelligent automation (IA) is already improving performance across financial services organizations. The first step in any successful intelligent automation program is to conduct a thorough opportunity assessment. Only then can you move on with confidence to start building an automation program.

Implementing intelligent automation requires the same approach as any modern technology delivery program; starting with the brainstorming and design phase by the team of developers. Once the design is finalized development can begin. Generally, an agile delivery approach with short sprints is the most effective way to build out the program.

Bring in the bots

Successive IA implementation programs can build on each other. It’s possible to look to the future, creating a “bot store” or “bot library” containing components that can be used in a variety of scenarios. For example, if a developer creates a bot that can interface with a specific application, a large part of that bot will be reusable in other builds to solve other business issues.

If an assessment and design phase has been completed properly, additional unidentified development steps, which results in delays and adds additional cost, can be avoided. It is also important to use an abstraction architecture process and technical design to allow the program to deliver minimum viable products in a phased approach.

Build efficiency can be enhanced when undertaken by experienced developers. Specialists skilled in using the necessary toolsets can build bots quickly. However, we recommend that any program that involves third-parties supporting the development phase should have a handover stage built into the program – to enable easy transfer of knowledge. Working closely with the operation and IT teams allows the development team to consider important factors such as business continuity, exception management and data security.

Furthermore, build success is also enabled when developers have easy access to client process SMEs, IT, HR and risk teams. Applying this delivery approach, means that the business, process and market experience are understood and reflected in the design and build stage.  

Choosing your delivery model

Financial services organizations have a choice of delivery models. It’s important to emphasize that with intelligent automation, it’s not a case of ‘one size fits all’. Firms are looking for flexible delivery models that work for them.

In the traditional joint automation delivery model, the organization is responsible for providing the IT infrastructure required for automation deployment, but works with a third-party to build, maintain and scale the program. This approach requires particularly strong governance and management to achieve successful execution.

In an alternative model, the third-party owns accountability for readiness, assessment, benefit realization, deployment and maintenance – managing end-to-end delivery – but the organization provides the IT infrastructure and support.

The third approach, known as “automation as a service”, sees a third-party providing a cloud-based IT infrastructure to build, deploy and maintain automation. The third-party manages end-to-end delivery – owning accountability for readiness, assessment, benefit realization, deployment and maintenance. A help desk can also be provided to give support to the organization’s internal users of automated processes. This model enables an accelerated return on investment due to the potential for rapid build, deployment and maintenance. Businesses are also freed up to focus on developing their own automation centers of excellence, developing capability and ultimately the potential to take over the running and maintenance of the IA solution in the future.

We’ve seen the “automation as a service” model embraced by organizations that want to achieve accountability and governance in a way that enables IA to achieve its maximum potential.

Managing IA

Successful IA builds in financial services are already addressing a range of end-to-end processes and business activities. For example, EY has been working on programs to help businesses build out their own automation assessment capabilities as well as establishing centers of excellence and control rooms to achieve correct management and governance.

In summary, a successful IA build can be helped by:

  • Choosing the delivery model that works best for the business
  • The right approach to automation design is key before you begin the build
  • Use experienced developers – but also enabling knowledge transfer to in-house personnel
  • Draw on a blended team of specialists who understand the industry, functional and process requirements, as well as IT
  • Developing a bot store to enable the rapid delivery of automation.

Summary

The success of an automation build depends on a number of factors. From the assessment phase to the teams involved, it’s important to consider these elements to optimize the build.

About this article

By

EY Global

Multidisciplinary professional services organization

Related topics AI Automation Technology