Will Artificial Intelligence Destroy Humanity
Cognitive Automation and the Disruption of Business Services
Legal Knowledge Management and the Rise of Artificial Intelligence
UC Health Explores Data Science to Help Improve Operational...
The Key to Effective AI is Not Technology
By Zorica Todorovic, Senior Vice President of Operations & Information Technology & CIO, Chubb Insurance Company of Canada
Technologies are just tools. The key to their impact on society and our organizations is how they are used and implemented. In my combined role of head of Operations and CIO for the Canadian subsidiaries of the largest global P&C Insurance Company, the key to our track record of successful project delivery lies in how effectively we use our technology toolsets to positively impact our operating efficiency, release of new products, and, most importantly, our customer experience.
The precursor to AI was ‘BPR’ (Business Process Reengineering). I would argue that this is still the most important part of getting your AI project delivered–both in terms of deployment and in achieving the anticipated ROI for these innovative projects.
Since 2002, by combining a deep business understanding of Operations with superior technical skills, we have been transforming our internal processes, increasing speed-to-market for new product releases, and evolving our interactions with customers. We started by mapping out the key business processes at a very granular level from first contact to product delivery. We did this detailed ‘As-Is’ process mapping for each transaction type and each product, completing an end-to-end analysis and documentation of our business operation.
This work was done jointly by IT and Operations, providing the ability for the IT staff to become intimately acquainted with our business operating model (BOM), but also to develop important relationships with our customers and users.
we have been transforming our internal processes, increasing speed-to-market for new product releases, and evolving our interactions with customers
The design phase for the platform was jointly executed together with IT, Operations, and business unit teams, which enabled us to build an end-to-end platform that, mirrored the workflows. From the initial delivery of the platform in 2002, we have focused on ongoing delivery of capabilities and products, with over 50 products and their custom workflows now being supported, these 16 years later.
The term ‘Agile’ was not as pervasive as it is now, but we call our process ‘Agile-ish’, as it uses the key tenets of minimum viable product, cross-functional teams, product owners, and an iterative, continuous improvement approach to capability delivery. Through this process, our AI platform has delivered on the business goals we set out to achieve. We developed rules engines that automated more than 60% of all processing tasks, which allowed for re-deployment of staff to knowledge-based work and significantly reduced our bottom line operating expenses. We were able to core our operation into two processing centers, and developed rules engines to drive work items to the lowest expense and highest capacity team members. Finally, this internal automation positioned us to deliver early digital capabilities to our customers, which we continue to build upon today. Equally important, these successful deliveries critically provided us with the discretionary funds to reinvest in larger technology initiatives, further enhancing our efficiencies both in IT and throughout the organization. The result: less than 40% of our budget is spent on Maintenance, and we are able to self-fund all of our initiatives using our BAU budget.
Another ongoing topic of concern for CIOs is the alignment to the business or having a ‘seat at the table’. A real problem in the IT industry is our tendency to focus on technology for its own sake, rather than on the business benefits the tools can bring. AI is just another one of those same opportunities. By first understanding the business and executing very deep process mapping, including partnering with the humans that do the work today, AI technologies truly do give us incredible power to transform and automate in ways that we have never had before. However, without this deep analytical and business focus, these technologies will pass us by as just another IT fad that fell short of delivering on its potential.
It is still a sad statistic that 70 percent of IT projects fail [source: 4 PM], while a shocking “17 percent of IT projects go so bad that they can threaten the very existence of the company” [source: McKinsey & Company in conjunction with the University of Oxford]. In my experience, the greater the hype and potential associated with a technology, the greater is the potential for failure. AI tools are extremely powerful and businesses in all industries today are building their future operating models on the promise of these efficiencies. Billions, even trillions of dollars are being invested. As business professionals first and CIOs second, we must do the hard work of keeping processes at the forefront of our AI projects. It’s not as good to slog through hundreds of hours of workflow review, and may not seem as ‘Agile’ to do this analysis and design work up front, but it will pave the way to truly Agile delivery of capabilities and, more importantly, to successful AI projects that actually deliver on their promises.