Norbert Monfort, Vice President, IT Transformation and Innovation, Assurant Global Technology
Those that have been in IT long enough know that while change is a constant, the rate of change is not. In fact, it has been accelerating for decades. So, how do we handle the onrush of technological advancements effectively? The key is to identify and focus on the technologies that are likely to bring your organization the most value. Technology that is not useful today could become critically important in the future. It is of utmost importance that there is continual awareness of how new technologies are evolving. This inherently means that organizations will choose to disregard some technologies and will gamble on others that they hope will bring in the highest return. In order to accomplish this, three basic principles should be followed: broad understanding, rapid testing/decisioning, and focused scaling.
Broad understanding means that we stay abreast of the evolution of new technologies, especially through the use cases being experimented or implemented. To this end, an insurance or financial services organization should not only use public information and research firm data but should also invest in InsureTechs and FinTechs related to the markets they serve if they have the venture capital resources. All of this helps keep a pulse on technologies to enable quicker decision making in future stages. As an example, an organization may be tracking and monitoring the evolution of Blockchain closely, but not make it a priority or area of focus until there is a compelling use case that brings value to the company and/or its customers.
An organization’s machine learning investment should remain within areas where exclusive data is available to solve specific business use cases and/or issues
Rapid testing/decisioning is the second principle. Once it is deemed that a technology may be useful, an organization should quickly test and decide on how to proceed. A key to accomplishing this is having the ability to create the necessary infrastructure and software installations in an automated fashion, such that IT resources are not needed. Critical to this principle is not being afraid to fail, especially if you fail fast and fully understand why the failure occurred. As an example, several years ago containerization using Docker/Kubernetes and machine learning were technologies that quickly became prevalent among many organizations in the insurance and financial services sectors. Initial experimentation within some companies were deemed failures, but for very different reasons.
Containerization in some organizations has been deemed to not be a good fit because its success is predicated on a large-scale commitment by both infrastructure teams as well as application development teams, and not all teams were willing to support it or fund it. Some companies skipped containerization and focused instead on utilizing cloud technologies such as FaaS (Functions as a Service) and PaaS (Platformas a Service) as these utilize containerization at scale butare supported by cloud providers that have already made the necessary large-scale commitment. In terms of Machine Learning, an organization must ensure that the right data is available for the use case being pursued. Understanding this from inception allows experimentation in areas where the data does exist, thereby resulting in successful rollouts of this technology.
Focused scaling is the final principle. When an organization is confident that a technology will bring value, then the focus must shift to scaling the solution. As an example, once it has been determined that machine learning will bring value to the company, an investment should be made in the platform that allows data scientists to develop models more quickly. Similarly, with RPA, once value is established, an organization should invest in a platform and a partner that brings scalability to future deployments. This scalability must also bring high utilization. Any areas where an organization does not have unique data that brings competitive advantage (i.e., in data security, HR, Finance, etc.) will be better supported via third party products or services. An organization’s machine learning investment should remain within areas where exclusive data is available to solve specific business use cases and/or issues.
While these principles may seem like common sense, developing the discipline to stick to them amidst competing internal forces pushing for a favorite new technology is a challenge that most organizations face. Still, these principles must be a starting place in order to bring discussions and investments back to a common foundation for decision-making. Only then can an organization effectively adjust to the ever-changing tech industry landscape.