RPA Health Check

Krista:
Thanks Sarah, for joining me today. How is your day going?

Sarah:
So far so good.

Krista:
You put on makeup this morning and I showered, which is a big day in the times we're having.

Sarah:
Yeah, I washed my hair last night. And I was like, "Wow. It looks so much better."

Krista:
So, Sarah you're one of our competency leads for RPA. And what I want to talk about first, about how many RPA bots would you say you've built?

Sarah:
At BP3, about 10? Or so? Yeah.

Krista:
And then, before that you were working on RPA as well?

Sarah:
Yeah, before that I couldn't tell you. Probably about the same number.

Krista:
And so, you've developed overall kind of best practices, and things that you follow when you build your bots. So, today we're talking about an RPA health check. What does it really mean to have a healthy bot environment?

Sarah:
You want to make sure that you have standards and practices in place, and that they're documented and repeatable. Sometimes you have to be flexible when it comes to developing bots, sometimes there are surprises, but you should always have processes in place to help as much as you can, to get bots into production. Because that's what matters.

Krista:
So, for anybody watching on YouTube, you can ask a question over in the chat for Sarah. Sarah, why don't you go ahead and takes us through the questions that you would ask in an RPA health check?

Sarah:
All right, sure. Let me share my screen and start the presentation. All right, can you see that?

Krista:
Yes we can.

Sarah:
All right. Where are my notes? So, first question; how is your automation workforce performing overall? That's pretty straightforward. How are things going, from a high level? Perhaps more importantly is two; are your bots delivering expected results? After going through all of the effort that it takes to discover, develop, test, and get a bot to production, we want to make sure the bots are doing what they're supposed to be doing, and that they're providing meaningful results.

Sarah:
Three; what is the general sentiment from human workers towards the bots? The last thing you want to do is to make someone's job more difficult by adding another confusing tool to the mix, like an unruly bot. We want to make sure the bots are providing value to the end users, and that the bots aren't difficult to work with. Fourth question; are there development best practices in place such as source control and code reviews? At first, it might seem like developing RPA projects is vastly different than developing traditional software projects. But, you can still apply the same methodologies and implement standards to ensure that quality bots are being produced.At BP3, we run our RPA projects using Agile development methodology, just like we run all of our development projects.

Sarah:
Five; are you creating reusable libraries and toolkits for common code? You definitely want to avoid duplicating development effort when you can. You don't want two different people working, or writing the same or similar code. Six; how long does it take you to answer the question, "Are all of my bots up?" Do you have visibility into the status of your bots? Seven; are you reporting on bot usage? Now, this can include knowing when the bots are running, what they're doing, what the volume of input is like, how often they're succeeding and failing, why they're failing, et cetera.

Sarah:
Number eight; how long does it take for a bot to go to production after testing is completed? Sometimes this can be a bottleneck in the RPA cycle. You want to make sure there's a defined, repeatable path to get a bot into production. Our leader Scott recently said that, "Go lives are always the most important things we do for our clients," so that's definitely something we want locked down. Nine; are you getting the most ROI out of your robotic infrastructure? Do you have the right licensing in place, the right amount of systems provision for your needs? Is your environment optimized? Bots can work 24/7. You want to make sure you're getting the most out of them.

Sarah:
And 10; are you actively improving your bots? I've developed a lot of bots in my time, and I have rarely seen a case where you develop a bot, push it into production, and set it and forget it. Business processes are always changing, and there's always opportunities to expand what the bots can do. With our Manage Robotics Program at BP3, we work on enhancing and fixing the bots that need it in a continuous improvement cycle, to make them better.

Krista:
Well, thanks Sarah. You can stop sharing your screen now, but we have a couple questions in our chat. The first one is, "What are some good practices in bot failure notification?"

Sarah:
That depends. So, you may want to send messages to the developers and the support team, in which case I would just send an email saying, "Hey. Something went wrong, find the logs here." However, if it's a business user, logs aren't going to mean anything to them, but you might want to let them know that, "Hey, your [inaudible 00:06:20] process that you're waiting on, something went wrong. Don't worry, support's on it." So, something more friendly and digestible that won't make them worry.

Krista:
Something especially to make sure that people know that it's being fixed.

Sarah:
Yes.

Krista:
Nice. I also liked this term that you used, unruly bots. And so, the next question is, "Is there a commonality you've seen in unruly bots that people can avoid?"

Sarah:
In general, you want a bot... Things are going to go wrong sometimes. That's just normal, what's going to happen. So, you should always plan for them. You should always, in your code ask, "Okay, what if it doesn't click this? What if it can't go here? What if it fails at this spot?" You want the bot to be able to fail gracefully. You want it, if it messes up, to not mess up in a catastrophic way. Or a way that requires a user to do a lot of manual work to get it back on its feet. So, unruly meaning it needs its hand held by humans, a little too much.

Krista:
Reminds me of Buzz Lightyear, falling with style.

Sarah:
Yes. Exactly.

Krista:
So you mentioned BP3's Managed Robotics service, for continuous improvement. What does that really look like in practice?

Sarah:
We use Rally to track our development work, we use it to track future work, features, enhancements, and defects. So, we are constantly having items tracked and worked on so that we can continue improving the bot. We have close contact with the clients, we have regular meetings with them, full transparency. And we're welcome to accepting feedback from users all the time, we want to make sure that we don't just say, "Here's your bot, and that's it, that's all we're going to do. This is all it can do." We want to make sure that if someone has an idea for something they want in the bot, we're like, "Yeah sure, we can try that. We could do it."

Krista:
So, shiny new bot and goodbye is not a winning strategy? Is that what you mean?

Sarah:
No, no.

Krista:
Well, thanks Sarah for sitting down with us today, and talking about an RPA health check. I think now we know a lot more about how to check the health of our bots.

Sarah:
Cool. This was fun!

Krista:
All right, well we'll see you again.

Sarah:
All right. Bye!