Many companies are curious about how they can improve business performance by using “intelligent” or “AI” technology to automate corporate processes and decision making.
The application of AI promises to be transformative, so as excitement around it grows many vendors have adopted its language. Unfortunately, this kind of technology is relatively new so many people use different and sometimes contradictory definitions of what “Intelligent Automation” and its various components is and isn’t.
This has created a lot of confusion.
These definitions around Intelligent Automation and related technology are the ones Exact AI uses internally and with our customers.
Artificial Intelligence (AI) is an umbrella term encompassing many different technologies and scientific disciplines. Conceptually, it is the simulation of human intelligence processes by machines, especially computer systems so they can perform human-like tasks.
Well-known applications of AI include expert systems, natural language processing, speech recognition and machine vision. AI programming focuses on cognitive skills such as learning from experience, reasoning and self-correction.
(I’ll be posting more about AI technologies and its scientific disciplines soon, including the various types of Machine Learning).
Robotic process automation (RPA) is generally agreed to describe systems for automating manual or repetitive processes that follow fixed rules. But it is not “AI”, rather its tools typically mimic the inputs a user makes in their software. For example, copying and pasting data or saving email attachments is a common RPA task.
Cognitive computing (CC) uses many AI technologies and disciplines to strive for natural, human-like interactions with a machine using models that learn, reason with goals, and interact naturally with users. Its biggest challenge is to emulate human thought processes in complex, ambiguous and uncertain situations.
Cognitive automation (CA) is software that uses multiple Artificial Intelligence (AI) technologies to emulate human cognitive abilities of learning and reasoning to automate processes that are information-intensive and require judgement-based processes and decisions. It is closely linked with both robotic process automation and cognitive computing.
Intelligent Automation / Intelligent Process Automation (IA or IPA) is a broad term often used by software consultants and vendors. Generally, it seems to encapsulate enterprise-wide process and workflow redesign to improve business performance using some or all of the above approaches and technologies. It includes both corporate-level and task-level workflow processes, and heavy use of business data and analytics for decision making.
One example is transforming procure-to-pay to automatically manage supplier trading from contract compliance to payment terms. This often involves both corporate-level and task-level workflow across multiple software platforms, departments and even countries.