Mining For Value With Process Intelligence

Prankur Sharma: Automation has become the mainstay of enterprise technology journeys today. And it really peaked during the pandemic period to epic proportions and within automation, Process Intelligence as we know it today, is one of the most exciting segments in the broader Automation space. It has emerged as a critical capability that leading Automation platforms are acquiring for and also a very exciting space from an investment standpoint, both from the enterprise side as well as both VCs and private equities going ahead and making a bet with their money.

However, as our analysis has revealed there are only 25% of global enterprises that are leveraging tools-based Process Intelligence platforms today and majority of those enterprises continue to rely on manual crowd sourcing techniques for discovery.

How can enterprises leverage Process Intelligence to supercharge the automation journey? What kind of efficiencies can Process Intelligence bring in and how has the broader Process Intelligence landscape evolved?

Hello and welcome to the all new episode of Zinnov podcast Hyper-Intelligent Automation series, and I am Prankur Sharma, Principal, Automation /AI at Zinnov. To discuss the nuances of Process Intelligence in depth, I have with Stephen Siciliano, VP, Power Automate at Microsoft, who is joining us for the second time on our HIA series.

Welcome, Stephen. It’s a pleasure to have you back on the show.

Stephen Siciliano: Great. Happy to be back myself.

Prankur Sharma: So let’s just dive right in. We’ve seen Microsoft has been making giant strides in the Automation space in the last few years. What have been the major product announcements from the last 12 months? It’d be great to get something said from your side.

Stephen Siciliano: Yeah, absolutely. So there have been a few big things that we’ve been working on. With the way that we think about automation, as we think about Power Automate as an intelligent Automation platform, that brings together a bunch of different Automation technologies under a single product. So one of those technologies is of course RPA, Desktop Automation on people’s PCs. And in the RPA space one of the big things that we’ve done recently is we have made a new offering available where Microsoft will manage and host the Virtual Machines that actually run those RPA bots on behalf of the customers. And the reason we did this is because we’ve heard from a lot of customers that the cost and the energy and the effort to maintain and run the VMs that run those RPA bots is very high.

And especially, over the past couple quarters with potential economic impacts and the macroeconomic conditions, a lot of organizations are looking at ways to save money. So by having Microsoft host the infrastructure, RPA infrastructure on behalf of the customer, this is a great way for organizations to save money by not having to spin up a bunch of VMs and have a bunch of extra cost.

So I’d say that’s one area that we’ve been looking at. Another one is in the Cloud Automation space. With Digital Process Automation. and here we’ve added dozens of connectors every single month. We now have over 800 connectors that you can connect to with Power Automate.

We’ve added new capabilities like the ability to have custom code as a part of your connector so you can actually write C# (C Sharp) code and now deploy that as a connector for the Cloud Automation that you’re running in a really scalable and resilient way.

Another part of this Intelligent Automation platform is of course intelligence. So we’ve been continuing to add and invest in new AI models as a part of AI Builder. New things like the ability to process paper documents that are unstructured, things like freeform documents and documents that have content that’s handwritten, so that way you can parse that in and build workflows on top of that.

And in the final area that we’re focused on, it’s of course Process Mining we talked about that just a couple minutes ago. But the big thing we did is back in March we acquired a company called Minit and I’m sure we’ll talk a little bit more about that in our conversation today. So those are the big areas that we say is part of our Intelligent Automation platform.

And the one other thing I’ll throw out there is across all of these, we think that governance and management are really really important too. So we’ve done a lot of things to bring more governance capabilities and management capabilities into the platform. Things like managed environments is one big thing that we’ve released recently that create a whole bunch of new management capabilities for controlling who can do what inside of environments as they’re deploying RPA and cloud automation in their organization.

Prankur Sharma: Great. I think a lot is going on for you, Stephen. I think it looks like… yeah, but I’ll just speak on the last point… on the whole Intelligence Process Mining space. You recently acquired Minit. You had earlier invested in FortressIQ which is another significant player in the category.

Tell us more about the Process Intelligence space. How do you think that confluence with Intelligent Automation is coming along in the broader sense as well as specifically for you at Microsoft?

Stephen Siciliano: So I think about the automation journey as a bit of a cycle. So at the beginning of your cycle, you need to be able to understand what are the challenges that people in the organization are facing.

And today a lot of that is very ad hoc. You know, somebody gets frustrated doing the same, boring, repetitive thing over and over, and so they go to, you know, the internet and see if they can find a tool that can help them out, and then they kind of stumble into Automation, they stumble into RPA or Cloud Automation.

But what’s really valuable about Process Mining and Task Mining is that can help organizations to identify the parts of processes that are taking a lot of time but that don’t require a lot of strategic thinking, don’t require a lot of manual input that are really those rote repetitive manual tasks.

So Power Automate’s process advisor is the capability that we had even prior to acquiring Minit and that has capabilities that will actually recommend where in a manual process you potentially want to automate something. So I’d say it starts at the beginning of the cycle and then once you’ve identified what to automate, you of course build that automation and you can use Power Automate to do that.

But the other thing that I think is really interesting is it kind of brings you back to the beginning because you can then use Process Mining to analyze the process after you’ve automated it. Because after you’ve automated it, you want to understand what is the potential ROI. How much better have I made the process? Are there still ways that I can improve the process? So that’s something that you can then use Process Mining for again and start that cycle over again.

So that’s kind of the way that I think about it and the capabilities that FortressIQ brought to the table, as well as what Minit brings to the table, I think are important building blocks in addition to what we as Microsoft have been developing for a while with Process Advisor.

Prankur Sharma: Got it. And I think we are overlooking at Process Intelligence as having these three components – Process Mining, Task Mining, and process Mapping. From what we understand, Minit is primarily focusing on Process Mining. So are you looking at augmenting this further with Task mining capabilities in the future?

Stephen Siciliano: Yeah, so what we have today inside of Power Automate is a Task Mining tool that’s actually based on the same recorder that we use for Power Automate’s desktop capabilities in RPA. So our focus on Task Mining is to continue to enhance and to develop that tool that we have built right into the product, leveraging that same recorder. And the nice thing about using that same recorder is many people may not know this, but Power Automate is actually part of Windows 11. So everybody that has Windows 11, in fact already has that recorder available to them if they so choose.

So it’s very broadly available. So we think that the Task Mining capabilities that we’ve developed in-house by leveraging that recorder from Power Automate will really help people for those specific tasks when they’re looking at a specific thing. How do I take these 10 steps and understand what’s going on for myself or for my colleagues that are doing that same thing?

And we think that’s a really good compliment to the capabilities that come from Minit in the Process Mining space, which is much broader about a process that involves multiple roles and multiple systems inside of the organization. Understood.

Prankur Sharma: And our analysis has revealed that today, like most of the companies are really at the fence of actually adopting Process Intelligence. In fact in our assessment, less than one-third of the companies that have started with Intelligent Automation are today using Process Intelligence technologies. Based on your understanding of the interactions with customers, what are the key reasons that you have observed for this and you see this changing in the next 12 to 24 months?

Stephen Siciliano: It’s a great question. One of the big challenges that customers have an adopting Process Mining is getting the data into whatever Process Mining tool that they have. And this is because the data is often in many, many different systems. And it’s not necessarily in formats or layouts or configurations that make it easy to directly port it into the Process Mining tool that they have. There’s also just basic things around permissions that sometimes can be challenging to get access to the data that your organization has. So we think that the data access and the data manipulation are one of those initial building or initial roadblocks that prevent.

Organizations are getting a lot of value out of Process Mining. And the unfortunate thing is you can’t really get a whole lot of value out of Process Mining until you’ve already imported the data. So it has this unfortunate characteristic of a very high initial investment is required before an organization can realize the ROI of understanding what their processes are doing.

And I think that this is an area that we’ve certainly been looking at Microsoft as well and the good news is we have a really robust ecosystem of connectivity. We have hundreds of different connectors that can pull data from a variety of different sources. As well as we have a technology called Power Query that makes it very easy for anybody to transform data to get it in the appropriate format that a Process Mining tool may need to leverage.

So these factors together we think can help organizations get to that value more quickly. But you’re right I talked to a lot of customer that’s a… they have another Process Mining tool, but they just haven’t gotten to that value yet because it’s taken them six months say to get the data into a place where it can actually be picked up by the tool and started to be analyzed.

Prankur Sharma: And I think for these customers specifically and audience who listening to this podcast show… you spoke about post-automation ROI is one of the benefits that enterprises can get a very good view on. So what other benefits do you believe Process Intelligence can drive all companies who are looking to deploy this technology at their companies?

Stephen Siciliano: Yeah. There’s a few different ones. One is optimization, so even ignore for a moment, understanding how much the process costs. Oftentimes there are just unnecessary steps that don’t even need to be there. It’s not… you don’t want to automate those steps. You want to delete those steps entirely from the process.

And in many cases, organizations find it’s those unnecessary steps that actually take the most amount of time. And until you have Process Mining as a tool, you don’t have that process map. So you may not even understand across the organization what all those stages in the process are. So simply just doing basic optimization, I would say is a big area.

Stephen Siciliano: Another area that’s really interesting is compliance. So making sure that everybody who goes to that process is actually following the required steps, this is the kind of the inverse of optimization is maybe there are essential steps that sometimes are being skipped and very rarely is this an intentional thing. It’s usually an accidental thing where people are not necessarily doing the thing because they didn’t have good documentation or they forgot, or they didn’t realize the importance of that. So that process compliance and optimization go hand in hand in many ways.

Another place where it can be useful is in benchmarking. So looking at best practices, looking at where people are doing that process best and then comparing how you’re doing it on one portion of the team or one part of the organization. And then of course it does roll up… I think the ROI because the ROI is super importing as well to understand what the cost of that process is and how you’re going to improve it.

Prankur Sharma: Got it. I think it definitely makes sense. How would you think of companies as they go about their Process Intelligence journey. Is it more like a one time activity that you implement the tool, you get the benefits out of it, you understand the process and optimize it, or is it more of an ongoing activity taking a more continuous improvement kind of approach? What do you suggest your customers do as they go about their process intelligence deployments?

Stephen Siciliano: Yeah, that’s a bit of an easy question. I would say it’s definitely a continuous thing to get the most value out of it. Of course, if you do it once, you’re going to get value. So I’m not going to say, don’t find a place, do it once, and then move on, if that’s all you can do.

But where you get the most value out of Process Mining and this whole area is by continuous improvement. And it goes back to what I was talking about earlier, where we see automation as a cycle where automation could be on one end of that cycle and the process optimization could be on the other end. And that just keeps going and going and going because you’ll find more and more places where you could potentially automate and improve that process, you run that intelligence again, see where those next challenges could be, see how you could further optimize and you just keep going and going.

So we definitely see it in the most successful organizations as being a continuous thing, and it’s a continuous thing that is it even just continuously looking at a single process but is growing and seeing what other processes, what other areas in the organization have that potential for optimization?

Once you’ve looked at one thing maybe you’ve optimized a specific process as much as you can, but there’s always adjacencies, there’s always related processes, inputs to that process, outputs from that process. So by looking at all of these things together, you can really transform the way that an organization works but you’re not going to get the most value if you just do it. It’s not a one time operation ideal.

Prankur Sharma: I think one of the other benefits that probably we didn’t touch upon earlier is also around sustainability. At least there’s a lot of chatter around it. I don’t know how much of it is actually coming to fruition. But yeah, the whole focus from a larger enterprise objective standpoint is to of course become more sustainable due to various factors and Process Intelligence solutions at least seem to be helping them further the cause.

Right. So what are your thoughts on this? Is this actually something that is.. how true is this and are you seeing customers who are using this from a sustainability standpoint as well?

Stephen Siciliano: It’s something that we’re seeing more and more of. As organizations become cognizant of their impact that they’re having, I think that, the way that we see it, kind of most interestingly, is in the context of supply chain. Understanding throughout the supply chain the total carbon footprint, those types of things. That’s something where we are seeing some interest by organizations and supply chain just generally is a very interesting place for Process Mining. You know, you’ve seen even over the past couple of years with Covid, there’s been a lot of supply chain challenges.

So understanding where those bottlenecks are and where your carbon footprint is as a part of your supply chain is something that organizations are very interested in. So it’s definitely something we’re seeing more and more of over time. I don’t think it’s now gone to the place where organizations are doing it enough. I’m sure organizations should do it even more going forward.

Prankur Sharma: Got it. And you spoke about supply chain now, and just picking up from one of the earlier responses, you spoke about compliance being one of the factors in terms of driving adoption. Are there certain industries due to these… I think certain… yeah driving factors… are you seeing more traction for certain industries or customer segment from a Process Intelligence adoption perspective?

Stephen Siciliano: Yeah, I’ll just speak to the experience we’ve seen. I think supply, finance as well as IT processes are probably the three biggest areas where we’ve observed it. Not to say that Process Mining is limited to those specific areas.

You know, there’s lots of cases in customer service and sales processes. There’s all types of different processes that are out there, but generally it’s going to be operations where we find the highest potential as well as the largest gap in understanding of what’s going on.

Usually for things like service or customer support those things have a little more modern systems, but for operations there can be a lot of older systems, a lot of systems that are difficult to get that visibility into. So that’s where we’ve seen the biggest need as well as in the IT space, because with IT there’s lots and lots of processes that touch on many different systems. So understanding how those things fit together has been important for organization.

Prankur Sharma: Got it. And what advice… as we are coming to the end of our conversation, what advice would you give to our audience, to companies who are looking to deploy Process Intelligence? They are somewhere on the fence and maybe they want to get started. So what would be your recommendation in terms of how they should get started and some of the best practices around it?

Stephen Siciliano: So if you haven’t done much yet, I would say don’t try to bite off a huge, massive project. Start relatively small. And it’s because of some of those challenges that I talked about earlier… getting all of the data into a single place, the permissions and the checklist you have to go through there. That can be challenging. So instead of trying to do the entire supply chain for your organization, start with a specific part of that you have access to that there’s potential for optimization in. Start small and even with a small process, you can still immediately realize value. And that’ll give you a sense of how it works, of what the building blocks are, how to get value out of it, and then you can always expand over time. You don’t need to boil the ocean on day one.

So that’s my general advice… is start small. Start with something you have access to. Figure out and get that ROI quickly. If you start small, you can get the ROI quickly, but then build from it. Don’t stop, keep that flywheel going of optimizing, optimizing, optimizing over time.

Prankur Sharma: Got it. So get some of those quick wins and then I think scale from there. This has been a wonderful conversations and I’m sure, I think our audience will find a lot of value and learning from this conversation. And we really discussed broke down this Process Intelligence space from understanding some of the key benefits, what are some of the challenges or more of apprehensions that some of the companies have, which may or may not be truly rightly placed as… and that just concerns which can bely addressed… how should companies get started on this journey, and then what are some of the truly magnificent benefits that they can derive out of it.

So thank you so much for taking out the time, Stephen, and it was wonderful to have you back on the show.

Stephen Siciliano: Thanks so much. I’m happy to be back again and I will look forward to our next conversation.