5 Lessons Learned from Intelligent Automation: Janette Fjeldstad, Circle K

The AI & Intelligent Automation Network asks key practitioners for five lessons learned from their intelligent automation journey.

Seth Adler

In a recent AIIA Network Podcast episode, Janette Fjeldstad reveals what has been afoot at Circle K and shares her top five lessons from intelligent automation...

1. Good and experienced RPA resources are difficult to find, also for external integrators, so best to make sure your integrator has a good team in place before starting to work with them. Otherwise you can end up as a training ground for their newest resources

  • Who’s training who? Get some viable case study information from your intelligent automation partner. Have them prove to you that they’ve got the skills and resources to accomplish your intelligent automation goals and deliver on and fulfill your intelligent automation strategy (you’ve got goals and strategy right?)
  • If you walk in blind, it’s your fault as your allowing your partner to create your goals and potentially even suggest your strategy. And at that point, if they’re testing on you- you ‘asked for it’ (by not telling them what you want and need).

2. Involving IT from the start when investigating potential processes is very important, as many times when we think we have found a good process IT tells us that they already have a way of automating it, it just hasn’t been done yet. To avoid any double automation work, IT needs to be in the loop.

  • Discussing projects through an executable timeline is perhaps the best way to get on the same page with IT
  • If they say they can do it, ask them, ‘within what timeline?’
  • This potentially showcases the fact that intelligent automation can be implemented quicker and cheaper than ‘usual.’
  • It’s at that point though that you should pull in IT and ensure they’re part of the process.
  • And it’s at this point (meaning right now) that you should 86 the us vs. them verbiage

3. Estimating the length of time it takes to document and develop a process on a bot, until it is ready for production, is difficult. Generally it takes longer than one expects.

  • “We can be up and running in 3 weeks!!”And I can cook you a meal in 2 minutes…that doesn’t mean it’s going to be good. Per Lesson 2, if we’re already way ahead of any timeline that the enterprise has already experienced, slow down. Do it right. Understand your process, understand how your technology plays with your process, understand how your team plays with both and then go.

4. It's important to start with several different types of processes as pilots in order to discover all potential challenges, especially technological issues.

  • Back to Lesson 1, do you know how this is going to go? Not necessarily. But if you’ve got goals and a plan in place, you can ‘test’ a few different processes with your partner and team to see what works and what needs work. From there you’ve got yourself a happy path roadmap.

5. It's also important to have a way of tracking benefit realization from robotics. Since many resources are repurposed rather than leaving the company, it can be less visible what the actual benefits realized are.

  • Intelligent Automation’s true and elusive benefits. Are you saving FTE? Great- welcome to the first baby step. You’ve taken incremental cost out of your enterprise. You’re nowhere near realizing the actual benefits to your organization. Back to your plan and your goals- do they involve digital transformation? Do they involve delivering on your corporate mission and vision in new and different ways?