How do you define intelligent automation?

I think of intelligent automation as a repetitive process with a feedback loop to ensure the task doesn’t vary too far from the central line of expected performance along with output on potential improvements to the overall process.

If you ask a person to stuff envelopes, they will do it and keep the letters within the parameters required by the postal service, but the stuffer may eventually suggest a more efficient process like sending a bulk email.

Setting aside the need for the stuffer to keep their job and whether hardcopies are being sent for tactile impact, more efficient processes are the goal.

Computer programs are great at replicating but creating value outside their existing parameters is more challenging.  Putting a human in the loop to reign in unacceptable levels of variation based upon  judgment is a good way to utilize both element’s strengths.

Intelligent automation, in essence, embodies the marriage of human ingenuity with technological prowess, transcending the boundaries of traditional automation by imbuing it with cognitive capabilities. It's not merely about streamlining processes but about infusing them with intelligence, enabling systems to learn, adapt, and evolve autonomously. In essence, it's the art of orchestrating technology to augment human potential, creating a symbiotic relationship where both humans and machines thrive in tandem, propelling organizations toward unprecedented levels of efficiency, innovation, and growth.

Intelligent automation can have three distinct pillars - workflow automation, advanced analytics and AI/ML (predictive AI, genAl and multimodel AI). Many enterprise use cases tend to fall under these that can elevate data-driven. intelligent automation 

I believe spatial (Apple Vision type) will be one of the key trends. Another trend we are working towards is a Conversational bot type framework with LLM and RAG.