I recently caught up with John Purrier, CTO at CA Automic, to discuss the affect artificial intelligence (AI) will have this year on business processes. John is Chief Technology Officer at CA Automic, a leader in business automation software (acquired by CA Technologies) where he is responsible for driving Automic’s automation strategy. Previously he was most recently CTO of CenturyLink Innovation Labs, and before that was strategic and technology leader at Cloud Foundry-based AppFog.
Daniel D. Gutierrez – Managing Editor, insideAI News
insideAI News: In what ways can artificial intelligence help power essential business processes?
John Purrier: There are two essential elements provided by AI assisted automation that will drive efficiencies, security, and performance in business processes. These are 1) intelligent decision processing, and 2) utilizing analytics and machine learning to optimize processes over time.
Intelligent decision processing can be thought of as the “recommendation engine” for machine-to-machine transactions, similar to how Netflix provides recommendations for machine to humans. When a business or IT process needs to “decide” a next step, or stage, in a workflow the AI system will present a set of choices, ordered by the highest probability. Today, these process decisions are hard coded or require manual input by a person.
By collection of data on “decision chosen” versus outcome and continually processing analytically, we create learning datasets for AI driven machine learning systems. Over time, this feedback back into the decision making process will create more and more optimal results for the business as the system “learns” to adjust the workflows based on business and IT environments and desired outcomes.
insideAI News: Which business processes is AI most likely to assist?
John Purrier: Datacenter, Cloud, and IT process automation are likely first candidates for taking advantage of AI assistance. Moving to a model that allows for self-healing and auto-remediation of issues through AI allows for so called “lights-out” operation of datacenters. This translates directly to cost savings to the business, and has the added benefits of producing repeatable, reliable, and provably secure automation processes.
Over time, AI will enhance business processes, taking inputs such as budget, SLA requirements, compliance, and audibility and feeding these into the intelligent decision making processes for business and IT.
insideAI News: What are the benefits of adding AI to business processes?
John Purrier: Automating business processes has the benefit of creating repeatable and reliable process workflows based on best practices. Adding AI to the mix enhances the automation with the ability to self-optimize over time, through predictive analytics and machine learning based algorithms.
insideAI News: How can AI-enabled improvements in business processes be measured?
John Purrier: As AI assisted systems become prevalent it is likely that there will be direct cost-savings through automation of tasks that humans have had to perform to date. For instance, IT professionals carrying pagers, in case there is a problem at 3am. Having self-healing systems it is likely that these sorts of problems can be solved by the AI automation systems without the need of human assistance.
There are also indirect costs savings as systems optimize over time. Performance and throughput goes up, there is less downtime due to human error, systems run 24/7 and remediate problems in real time, and AI enhanced security protocols lessen exposure to hacks and denial of service type attacks.
insideAI News: How can IT win over AI skeptics within their organization?
John Purrier: AI needs to be introduced in a non-threatening manner. For a lot of people, AI is akin to magic or is what powers the Terminator type robots. AI enhanced systems need to be transparent and auditable, and introduced incrementally into the organization. Seeing is believing, results will help win over the skeptics. AI systems should be positioned as enhancements to existing business and IT processes, allowing machines to help optimize existing best practices no displace what has worked so far.
insideAI News: How can a business get started with AI for business processes?
John Purrier: AI has matured to the point that it is practical to use in many software driven scenarios. It has been stated by Google, Microsoft, etc. that within 5 years all software will have elements of AI embedded. Businesses should be looking to solve their business and IT process needs, and then look to vendors that are innovating within these solutions with AI and machine learning systems.