Courtesy of Alberta Health Services' Jesse Tutt below is a transcript of his speaking session on 'Implementing RPA at Canada's Largest Organization and Freeing up 10 FTE of Capacity in 180 days' to Build a Thriving Enterprise that took place at the RPA & Intelligent Automation Live Virtual Conference.
The Intelligent Automation team, at Alberta Health Services, that's Canada's largest organization.
And I thought it was the trucker's on the border was the largest organization. Anyway, HS as an IT team of over 2700 is working on a one-point-five billion-dollar IT program. Think about that, IT is only part of the health system, right? So that's a pretty big, pretty big deal. And I think a leader with 20 years of experience, his background is focused on intelligent automation, including robotic process automation, totally relevant for today's software development, ETL processes, artificial intelligence, and analytics.
And Jesse excels as an entrepreneur, founding a software development, virtual tour, mattress, and search engine optimization, SEO company. I love it. So, what Jessie represents, for us, is a very specific perspective, which is great, around the health care space. Jesse, I will turn it over to you.
Tell us about the wonderful achievements you've got.
Thanks, Chris. I'm just gonna share my slide here quickly.
All right, well thanks everyone for joining me.
Look forward to sharing the story about Brunelle Services, specifically in the context of the automation work that we've been doing.
We have been working on some really exciting things and look forward to sharing those with you. A bit of background on Alberta Health Services. We're basically Albert Canada's largest organization. We have 180,000 users of our systems, 127,000 staff.
And the others are, you, know, for the most part fully funded companies subsidiaries the, we support four point five million Albertans.
We are basically in the top five most integrated health care systems in the world, and just in terms of the geography of it is around five times as large as England. And about 1.5 times as large as California. So, IT, obviously, a big, big IT team. And from a clinical perspective, we're one of the larger organizations for those in the states for kind of comparable size to the Kaiser Permanente's of the world.
Personally, as Chris shared, my background in IT has been in IT for 21 years.
Tons of experience on the, on the automation, you know, ETL processes, analytical, data warehousing, technical architecture, and then, most, most recently, or not most recently, but marketing side, and, of course, the debt. So, gotta get that.
Basically, slinging sling diapers as well.
The summary is, at a high level, in 2007, we ended up starting to use robotic process automation.
And so we were an early adopter. We have probably about 150 digital workers in Boston Workstation today.
Simply running reports, and there are a few other areas that they help with, but a lot of reports running in. 20, 21 passwords, quite a few years, we jumped into releasing an RFP and awarded that to Blue Prism.
And today, we've gotten basically from 0 to 91 digital workers and freed up around 17 FTE 17.3, I think as of this morning.
So from a vision perspective, we're heavily focused on returning capacity back to the business. So as you can imagine, you know, with the Cove it and then the financial constraints of of ..., our organization is centered on trying to free up capacity, improving quality, you know, helping the patients.
And naturally, our customers in the organization, where there's, there's a lot of different stakeholders through the development of digital workforce to augment or actual, you know, staff workforce.
The principles are, are heavily focused on Digital first. So we've got a and I would add a bias toward action.
So, you know, when we're looking at automation, we look at all the different options. And, you know, oftentimes, there's a capability in system. There's a capability through RPA, there's a capability through software development. And we tend to side with the one that we can free up capacity. The fastest communication is really key. I've got another future slide on, the on change management.
But the whole principle of ensuring that we have a very clear communication model, it is necessary for any RPA program, arguably more important than the techie stuff, to make sure that you are managing all of the resistance and the concerns that can come across, that can come from the business when you're implementing our RPA.
one of the first steps, I think, in any RPA journey, is, is to to get, get all the executive sponsors together and decide what's, what's the vision. What does success look like in this year, or in five years? And so, you know, in our case, we took a look at all the options. And we initially, were focused on cost savings, which certainly has some benefit. And the benefit is, definitely, you know, the CFOs. And the people that are sponsoring the program, can take that and actually offset some of the cost of the investment in the automation seaweed, which is natural.
Unfortunately, when you're only centered on on cost savings, you miss a lot of the opportunities that can come from cost avoidance, which innately always, almost always turns into cost savings because you're avoiding costs. They don't need to be invested in the future. So, we've shifted focus, We've gone, and now, now, we're looking at total total savings, which is super exciting.
It's allowed us to actually invest in a lot of automations that are freeing up capacity, which are making a really big difference.
From a structure perspective, there are certainly a lot of them that are out there. You know, centralized, federated. Naturally we've, we've gone with centralized model, from a governance structure perspective. You can see, we've got the executive sponsors, at the very top.
And, and frankly, this is your number one success criteria, for any automation seaweed. So we were fortunate, we've actually landed two executive sponsors that are direct direct reports to the CEO, which in our organization is a big deal, because a lot of our staff, we have up to eight management layers to the CEO.
It's a very large organization, and so to have that kind of sponsorship is excellent. In our automation Steering Committee, we have the heads of almost all the corporate functions, which is our focus initially. So we've got the head of IT, the head of HR, that of supply chain management and then the head of staffing.
You know, which is, which is great, and Finance as well, the automation CUI is staff. That's, that's my team.
We work closely with our vendor, We do have a program office that we do reporting to or provide reporting to, I should say.
Working, you know, at the operate more of the operational level, each department, the department head is assigned as the executive sponsor for the automation work in their portfolio.
They assign an RP champion, who is accountable really for it working with a management team, collecting the automation opportunities and and kind of making sure to keep an eye on the change management aspects of supporting communication to the departments, supporting that the department managers, to leaders and then naturally down to the, to the staff.
Whenever we're implementing new automations that impact them, we want to make sure that we communicate clearly So that they know the upcoming automation and how it will impact them.
So you can kind of, you know, what I'm trying to to get at is change management is really is really key and it's layered.
From a fusion team perspective, we have our Automation ... project manager who tracks all of the different automations.
So we've got over 50 different projects that are actually chartered in our CA PPM tool.
We have, business, process me, it was assigned to each automation who knows the process the best. So it's, it's, it's really important, I think, to layer on that project management for tracking of tasks to make sure work gets done.
Then bringing in the business. Who knows the process the best.
And then the RPA developer, of course, to develop the automation.
The next discussion, I think, is, what is the scope of the CUI, what is the scope of the type of work that goes through here?
And so naturally, we're still having discussions on whether, you know, in system automation is encoded in scope.
And currently, we're trying to, to, to, to remove in system automation.
But the, the lot of Teams focused strictly on RPA.
And I think the benefit is, it's easier, you, got, it, got a smaller team.
But the cons are, RPA isn't always the best case. Now, you know, maybe there's an API that you could use. Which, frankly, you can use in RPA, Alexander.
But maybe, you know, there's a website or something like that, you know, that you want to, you know, front end the systems.
The pointers, you know, when we're looking at an end to end process, specific tasks, are great candidates for RPA, and maybe another couple of steps. and up in the process.
You need a website, or some other mechanism, and so looking at it holistically and actually picking the best technology for each of the different steps in the process, is, how we're approaching it, The, it, I think, it allows you to support more use cases. The downside is, there's a lot more stakeholders, and so, you know, working through the, the, the political nature of overlapping services in the context of IT, or, and, or potentially even business areas, it, there's a cost.
But, in our, in, our, for our team, it's, it's had a lot of benefit from a pipeline perspective.
Collecting the automation opportunities is, is key. As I mentioned, the RPA champions or the Automation Champions are assigned to work in their business area to collect the ideas in their own business or department. And so we've got a really cool intake form, which we've leveraged a lot of the work that and they tested their intake form, which is publicly available.
And that upended a bit more of the, the logic, which allows the business to actually estimate the cost of automations, without even talking to us. So my team is not involved in costing of automations until later stages.
So at the ideation stage of the business, it's literally, the managers and the leads and the staff, are met, have frequent meetings, with the RPA champion. The RPA champion basically collects these ideas in a single Excel document, that has one column for automation.
We collected around 20 different points of data on each automation. And the output is a calculated to return on investment timeline.
So, you can see, that this, we're going to have, you know, the cost, this automation will be paid off within six months, or eight months, that. Secondly, the one-year benefit, as well as, the ongoing benefit annually, after that.
So, those three measures are key prioritization mechanisms for the business to prioritize their automations.
The our steering committee takes luck, prioritize them further, and if the business is OK and wants to push an automation forward, and if the steering committee wants us to work on it, we begin work on it.
And so we then go through, we work with the business to clearly articulate the exact process that they want us to automate in the form of a PDD, which is the process definition document.
Process definition document is is for the most part, largely automated in itself through the use of a tool called Captivate.
And so which is similar to step recorder, but it allows a lot more complex recording of processes and branching, which is integral to any automation because you don't want to know the, the, the 80% process. You want to know other options and the standard deviation, the deviations in the process. So we basically take all the PDD stuff, There's, there's further privatization of the Scope Changes, too much. And then, as time progresses, we work move into the development stage, design the automation and the STD, and then actually begin development.
So, from a collecting perspective, it's, it's really important that it's easy.
And, in our case, we've built this intake form, which, if anybody is interested, happy to share it, which calculates all the development costs and is a nice and easy method for the business to collect and prioritize their automations.
Governance committee prioritization. Now, there are many options, And just having worked a lot in the strategic side, in the technology stack and in other areas.
You know, I've seen elaborate models where you've got, you know, kinda 20 different columns, and in an Excel sheet, and you're, you're recording the strategic alignment, and, you know, payback period. And, you know, how fast are we going to get a return on investment and other other key measures.
In our case, we don't use any of them.
The business basically prioritizes their own automations.
And what that looks like is, for example, in finance, they probably have at least 50 plus automation opportunities that they've identified.
They've submitted, you know, maybe 5 to 10 and each of those automation opportunities then are reviewed and there's a presentation.
The Steering committee has either a thumbs up or a thumbs down.
At the point that it comes to my team, we handle them or in FIFO now, a lot of people might say, oh that, that's too risky And you know, you're not focused on return on investment payback period, and that's true.
But what we've seen is by giving every department a piece of the action, they're all happy.
And the second benefit is that sometimes you just simply need to do an automation with a poor return on investment to get buy in and to, to get people excited about the opportunity for automation and to instill the notion that these, these automations have a benefit.
So once we implement these automations and then gain their trust, then then it's a, we typically see a spurt of activity, and submissions.
From an operating model, there centralized and decentralized, from an automation seaweed perspective.
In our case, we are very much focused on centralized and why, because every organization that I've met, has started with centralized. And so, you know.
I think it's actually, with the exception of one, I did meet 1, 2 days ago, that was a large organization, government organization that was, was doing decentralized, Now, there's, there's pros and cons. And I think it really does depend on the organization.
But, I guess, the point is, is that RPA is a very technical competency.
You'll want to make sure that the infrastructure is set up to be completely resilient.
And that, that there's a cost to that.
And setting up, too, you know, environments cost twice as much.
Secondly, having resources that you can backfill and like in the, you know, calling in the evenings and things like this means that for most organizations, you want to make sure you've got, you know, 2 or 3 people that are capable of supporting things in the evenings or in the event that you rent tissues.
Ramping up. In our case, we naturally looked to external resources, you know, heavily in the beginning, and as time has progressed, we've kept those numbers.
Probably a little bit smarter than what we started with.
We've started to augment with students, and end onboarded additional resources in our team.
And, so, we, we now have a nice mix, in my opinion, of, staff, students, and external resources, and shifted a lot of the, the, the external purchase services from development of RPA automations to supporting us in, you know, reviewing or doing quality assurance.
Excuse me, I'm automating. Or not automating. But, reviewing our operating model, and which we refer to as our Automation seal, which I'll speak to shortly. And then taking the students to, to do a, lot of our automation work in. Designing the objects.
And then naturally moving our staff to be more of a senior role and helping to architect automations, that May cross, you know, code that may be functional across 2 or 3 automations and making sure that we're designing things in a very, you know, re-usable manner.
The Blue Prism, Robotic operating model, is, where are we started?
Almost all of our documents, which you've got probably about 20 different documents in total, but one main one that covers the end to end process, the foundations for a lot of those concepts came from the room. And I would definitely suggest checking it out, it's free, it's a, it's a compilation of around 100 different robotic operating model documents that cover the domain areas that are on the screen.
From a project delivery approach, we naturally start with collecting ideas, Like I mentioned, through the RPA champions, You know, the, I think it's, it's, it's really key to push a lot of that intake and discovery to the business. They know their business the best.
And having a person embedded in their team who knows their business, that can meet with our managers and their leaders, is, is something, I think that, that there's, there's benefit in having the business do it.
Because oftentimes, the automation developers don't have the business competency and they don't have the connections, it's just, it's been a, it's one of the one thing that's worked well for us.
After we collect the ideas, that's the business, prioritize them.
The them, they go to the automation steering committee and then we worked through the analysis and design stage document.
The PTT like we mentioned, are like I mentioned and then go to you know proceed basically with the design of the automation and the implementation.
one of the key differences, I think, that in areas that we are focused is process mining, which are we'll go through a slide or two on that in the, in the future.
Process documentation is, is obviously extremely important because it's important that we have a repeatable process.
That's, that's proven.
And so, my, my whole staff are trained on what we call the Automation and ...
way, which details all of our roles, as nice clear racy as the the entire NM process, not just for the project side, but for the operations. And covers everything in detail.
This is a sample, basically a table of contents from a couple of reversion versions back. But the point is, is that anything you need to know on in the context of the automation CWA or the automation CD processes is covered in this document. And there are links to supporting documents in some cases, but but it's not needed this covers at all.
The business process documentation piece takes a heck of a lot of effort.
If you're going to do it manually and create a Visio diagram, it can literally take months, and especially for larger teams.
So, we work with some teams that are one team specifically that has 300 individuals in it, and they have 10 offices and 26 teams.
And, as you can imagine, with such a large organization, some of the teams don't do things exactly the same way as the other teams.
So, this, the ability to actually capture the business process, map it out, and add branching logic, and add comments and taking screenshots, and then exporting that work to a Visio diagram and, more importantly, to the Blue Prism Software so that we can actually draft our automations adds a lot of value. Once again, it's free. So you can check that out.
But we found it to be extremely valuable and much more.
No effective, then, step Microsoft Step Recorder or other tools that are out there.
Our use cases, so far, we are sending out a ton of letters. So, we send out offer letters in our HR system. We send out regret letters.
If a person didn't get a job, we sought out cancelation letters or completing tasks in our new hire process that speeds up the delivery of our, allows us to hire people more quickly.
Position changes for it, we're handling a lot of that now with the automation work that we're doing terminations, many of the steps in the termination process are automated.
We're forklift data from one system to another and the finance team.
You know, right now whenever you complete your online training in the on the for the finance system, the automation will detect that training that's been complete and automatically provision access CP slash M which is our supply chain management team. We're flagging issues, you know, from an auditing perspective and loading data from one system.
In our case, Epic to Oracle Financials, staff scheduling, we're helping with, you know, our shift in our scheduling processes.
From an audit perspective, reporting perspective, deer perspective, and then, naturally operationally canceling and re broadcasting shifts.
On the IT side, we're just kinda getting kicked off with a lot of the IT work right now.
But, our primary focus is on supporting the IT Access and the desktop team, and I would add the Service Desk, as well, which are, kind of, the tripod of, of demand in, in most IT organizations.
Since around 50% of tickets, On the request side are typically IT access related, and more than 50% of the incidents are actually done and managed by the Service Desk.
Use cases, I think, whenever you're presenting RPA, a lot of eyes are going to get cost over. And so working through and presenting use cases in the, in the viewpoint and in the background of the business area that you're meeting with is, is really important.
So, this, this deck, or in this this link, which you can work, well, It should be in the handouts, allows you to, you know, pick and choose specific use cases when you're meeting with finance, or when you're meeting with supply chain management or contact center. So, there's, there's, that's my collection.
From our privacy and security perspective, naturally, RPA is new and, you know, right, when you're kicking off a new program, there's going to be a lot of people that, you know, raise concerns about. Automation.
You know, digital workers processing a chart content or maybe, you know, working in a clinical system are or how things are logged in and the access that's related to them.
So, from our perspective, we, we engage privacy extremely early and our information risk management team reviewed the whole plan.
And the consensus is is that, for the most part, our automations are doing the same work that humans are doing today.
And these humans are following a process that's been audited and so by starting the conversation like that, it does kind of at least, bring down a few of the barriers.
Secondly, we follow whatever access process they cuban's follow.
So when we started, there was quite a few different discussions around, hey, you know, should we follow it. If for, for digital worker needed, access to a system. So we have a dedicated, you know, RPA access process. Just to get the RPA accounts access to the enterprise systems. And we're right from the get go, we've always said hey, no.
You haven't been audited proven process. For humans, we're going to follow the same process, End up that, that's really allowed us to speed up access Provisioning. Which initially took, in the case of some of the earlier systems, Up to 4 to 6 weeks, to get access to a system. And now, you know, we often get access within the same day, maybe 2 or 3 days. So, it's made a big difference. We also have a standard naming convention for Oliver or accounts so that you can easily see that they are doing RPA.
Number two, in the actual system, you can see that the, the manager of that account is linked to my team, so that we can actually, the, the IT groups can actually get support foster.
Ostia is, is extremely important. We know that.
The success of any program, how to start, you know, certainly, I think, you know, finding visible sponsors is probably the number one.
Number two is acknowledging and planning for anxiety and resistance.
So, a lot of these, different, you know, staff, when they start hearing about bots, and you know they immediately jumped Terminator movies and, And explaining to them that, you know, these, these digital workers don't, no, they're not autonomous. They can figure things out themselves, You have to train them in the exact same way that people get trained.
You have to provision a computer just like a staff and embedding the notion that these teams are moving to both digital and stuff, resources, and that they're all basically part of the same team, and they're passing the baton of work back and forth.
I think, helps, helps a lot.
Secondly, mitigating risk product, you know, concerns about job security, and the impact on on, on work, and, you know, if your work changes, we're gonna work with here.
We're going to provide training and make sure that you know, that, get the work that you you hate. I love in the earlier presentation, you know, if you If you hate it, automated, I think that's a great tagline.
And I think it allows us to work through those those notions that are pre-conceived ideas around automation, people fear, proving the value. So one of the, the naturally, with an analytic political background and data warehousing, we ingest the records from our Blue Prism system into a reporting database.
And then we present that data in a dashboard that's updated every 15 minutes.
For every single stakeholder in the organization, you can actually see the exact automations, how many minutes that it takes that each queue item saves the organization.
If it's successfully processed, Excuse me. How many minutes of time that the business exceptions, which are the cases where the automation has been trained not to do, to, to handle, a specific queue item, which sometimes save time as well.
How many minutes, each of the, if you take the cue items that were successful times 10 minutes, and the, you know, the business exceptions times in zero cases for the top example.
The total amount of minutes saved across the organization is 3000 hours. And then, if we look at the time, that was saved to date, if you translate that into an FTE. It's 1.69, and if you extrapolate that out to the year, it's 2.31.
So, by having that visibility, and a dashboard that's update updated in near real-time, you, you prove your value. And it's a, it's a, you know, literally, I can forklift this our detailed dashboard.
Directly into our, into our, you know, any type of senior presentations or presentations and I'm working on to show that, hey, we, we don't have just, you know, theoretical value. We have proven value.
And I'm sure there's, there's going to be questions around. You know, how do we translate this financially And certainly, we have the capability. We do not have the salaries of each of the different automations on the average salaries for each of the automations. So, we can try and translate those into actual financials as well.
But for the public dashboards, which is, is a public dashboard, we don't present that data architecture.
So we have done a lot of work on this, and so, you know, when you're implementing digital workers, you've got choices. And really, you've got, you know, three depending on the, on the solution that you're implementing.
You have either virtual servers, which, you know, are costly, um, get virtual desktops, which are around 39% lower cost, and then you've got, you know, basically, ..., which is the Remote Desktop Services. Anyway, it's, it's a it's kind of Microsoft's Citrix technology, which we are looking into. And it has actually a significantly lower cost.
But currently, we're on servers and we're around 60% migrated now, I think, to desktops.
So we've kind of jumped from 1 to 2, and we're looking at three as a potential for the future.
Digital workers, meantime off, so when you're planning your your, your, especially at the business case level, and, in our case, the costing that's actually embedded in our intake harm. We don't plan for automations to work 20%.
You know, the digital workers, they can't run 24, 70. You've got to patch them, you have to upgrade them up to the enterprise systems, need to be upgraded.
The database has to be upgraded, and so my point is just to make sure that you've got a little bit of leeway to give yourself room that if the digital workers aren't available for a period that you can rush, you've got that extra capacity to run multiple digital workers at the same time, for a automation, that's parody the technology assessment. So we do go through very elaborate process of understanding each of the detailed steps in the process. From the PTT Then we have meetings with our enterprise architects are IT service owners or application owners.
Our business owners, the privacy team, depending on if, you know, there's privilege data, whether it's clinical or employee data.
And we worked through, for each of the different steps in the process, what is the best technology to automate that specific?
Now, this is a Southern Company Company diagram, which I like, which is on Gartner.
If you, for those that have a Gartner account, they the ability to actually go through each of the different steps in the automation and discern, is this a remote, robotic desktop automation?
You know, basically, for any type of, you know, live automation that you want to support, your end users, potentially, your service desks, you know, maybe a high administrative role, Is there, is there an opportunity to use robotic process automation?
For anything that needs to be scheduled, you know, I think there's that. That's typically the primary technology, the API now, naturally robotic process automation can, can use APIs, But the point is, you know, maybe there's an opportunity to use infrastructure automation tools, then, which we have in our organization.
Secondly, is the ETL processes in integration tools, like, you know, cloverleaf and SQL Server integration Services, which we, we do use SQL Server integration services extensively to ingest data. We ingest data from probably 300 different enterprise systems right now.
The digital assistance is one area that we're heavily looking into right now.
So, we do have ServiceNow for our IT and HR ticketing system, which has a really great chatbot capability, We're also looking into a couple others.
But the point is, I think having the ability for the end users to, to, you know, communicate in a way that they want to with the automations, the RPA automations. That's the best model. So, we, we have, you know, chatbot capabilities. We have automations that are triggered through e-mail.
We have part of our, sorry, we're working on automations that are triggered by e-mail.
We have automations that are triggered or without weren't, intending to trigger from web form replacements. We're no tools, whether that's interact, or whether it's a custom made the website.
Then we have automations, I can get triggered from files. So, some and are. We have some automation set.
The end users will literally the business will drop files into a folder.
The automation works and it moves that fall to the completed folder. And then lastly is the ... Cog cognitive computing piece, which is really that AI NLP ML piece.
And so, the, each of the different tools, whether you're using any of the larger tools, they all have that capability to use AI or integrate with using an API, with no Google Cognitive Services, or Microsoft, any of the different host, IBM Watson, any of the, The Online solutions and stuff.
The point is, is that RPA is, it's an entire stack of solutions that you can use to automate.
And, above that, there's other technologies that you can use to help feed data into the RPA automation.
But, looking at it holistically, allows you to, to support a lot more use cases and hammer out automations faster.
The identification of automations and so, you know, when we're going through automations a lot of teams, not a lot of teams, I should say some teams have challenges coming up with ideas, which.
Is a, is a valid, no concern.
You know, it could be corporate culture, it could be the fact that they're not technical. Or the fact that they've been doing a process for a long time. And they just don't have that viewpoint because of potentially, you know, winders, other ways that they could do it. Or a digital worker could even do it.
So training them on what ARPI automation or RPA can actually offer, You know, we'll get you a long ways down the path.
But for some, they need, they need help.
Now, the second concern is, is that, next, actually, at the root of every single business case for RPA, is the task duration. So if you take out automation, that's no 100,000 documents that are processed per year, And each one on average takes a minute of human effort to actually automate.
You're looking at 100,000, you know, hours, minutes, whatever the time, duration of the of the task is.
Now, when you're looking at the benefit, typically, we extrapolate that number from the one unit, out to a year, out to five years.
So it's extrapolated many times, which leads to an accuracy, which is why the task duration is, is an integral amount of it is integral to any business case.
So, typically, for our automation, as we work through, we use capture and we actually record the duration of the automations at least three times to give us a sense of, like, is this an accurate number?
But, one opportunity that we're looking into right now is process mining.
Now, process modeling helps an organization, at least in principle, and we're still working through this two ways.
Number one, that allows you to, to validate the duration of the task, which is the key, no measure for the, in the business case.
Secondly, it allows you to try and potentially identified new automation opportunities, and teams that don't have a great sense of that.
The way that it works is effectively by scanning log files. So, in our case, we're looking at ServiceNow, initially.
and so, we're ingesting the log files from ServiceNow literally as we speak today and tomorrow, hopefully. But we're working with the application teams to get those those extracts. They load that into different software tools. They're slowness.
There's avi top timeline.
Then, the data literally gets run through a machine learning algorithm. Which is the Middle Step.
And all of a sudden you see a view of peoples of the events.
And those events get rolled up into processes.
So, you can actually start to see, a, an understood, and understand 90% of the tickets are, you know, 100% start here, you know, 80% flow through to this next step.
And then a portion of the ticket's actually come back, Or maybe it's the portion of the invoices. Or maybe it's a portion of the patients.
So, process finding, I think, has really good, great promise. We are looking into it. And try to understand, not just the process of the person.
The one person that we picked, to record their, you know, through step recorder, or through capture, or through whatever tool that your organization uses.
But, to get a sense of, across all of people, in all departments, they use the system.
How does, how do they actually process things, and, more broadly, you can actually ingest the log files, not just from one system, but for multiples. So, you can see kind of an end to end view of the different items that have occurred in the process.
Task winding is the second technology, which I think is, is worth considering.
Most, process, planning solutions offer task modeling, process finding, is looking at the log files.
Task mining is for pulling the exact same process, But, looking at the recording, they're looking at recordings of the actions of stuff on there, that, the work that they do on their screen.
So, the machine learning goes through same kind of methodology, looks at all the different screens that recorded, then naturally collects the steps, identifies the steps, rolls them into processes, and the bumps. Those processes, allows you to do analytical work on the processes, to figure out which ones are optimal for automation.
They recordings on the screen, I should add. Most tools allow you to exclude things. Not just include, so, you know, and include everything is where, where I was leading.
So, we can say, Hey, you know, we don't want to include Facebook Messenger. We don't want to include Gmail, or, you know, anything, anything else that's private.
So, you've got that, that capability to record. The work that people are doing in a, in a, in a bit more of a, of a non intrusive manner.
The machine learning rolls it into processes, and then you can use all of this data to allow you to actually see, across multiple users, across multiple teams.
The, the precise steps that work has, is being done, compared to process mining, which is more at the system level, and a little bit less detail. So, most organizations, they start with process binding, to kind of delve into big picture into the areas, and then use task mining to drill down into the detailed tasks level view.
So, from next steps, my team is, at least in the short period of time, we are zeroing right into IT.
So, we started off with HR, extended into finance and CPS, um, and naturally the staff scheduling teams, our next area of focus is IT.
And so, you know, nationally, with, you know, 25, 2700 people, we have, there are certainly opportunities and the onboarding of process mining on the ServiceNow tool.
And we're hoping to use and approach IT with a bit more of a data driven approach.
Compared to just working through and actually collecting the automation opportunities to increase the, the return on investment from the automation work. and secondly, validate the task durations like I mentioned.
We're also input, frankly, we've already implemented decipher, which is the Blue Prism OCR tool.
So we are working on ingesting financial firms that are in PDF form, extracting the unstructured data like the blob of text on them and then loading them into our Oracle financial system.
So, having that capability to do OCR on PDFs is is another tool in the toolkit, which is kinda part of that cognitive function That, I mentioned The AI capabilities next up we are going to be implementing interact with which is a web form replacement tool. You know as I mentioned, you know having the ability for end users to be able to interact with the automations or the digital workers.
The way that they want to interact has the capability to trigger the to simply have a form on the on the on the screen.
Similar to all of our existing PDF, you know? forms that you fill out every day.
And then triggering the automation to, to trigger the second that the person click submit at the bottom of the form. So, allowing us to have that, you know, web form replacement, In our, in our organization, we have literally thousands, if not tens of thousands of forms.
And so, there's great opportunity to take those forms and have the work done in the automations.
Lastly, next, you know, 4 to 6 months, we are already kicking off meetings to start to look at process modeling of Oracle financials.
And after we get that done, certainly, the next intention is to start looking at Epic, which is our electronic medical medical record system.
Which is used by around 40,000 people so far. And will be used by over 100,000 in the coming years as we migrate people to it.
So, having an understanding of the the processes that people do on the court corporate side, and on the clinical side, is kind of our approach to two spring, and identifying high opportunity.
Are the automation opportunities that have a high return on investment, and kind of focusing our efforts on those.
And that's, that's it, Many questions.
Just, if you could please leave that screen, I just for a second just, So if anybody wants to write that down, because that's what I wanted to suggest, anybody can capture that screen capture, or you can get the information later after the session is over. We've only got a couple of, a couple of minutes left. We're going to end. We were supposed to end in about one minute, So, anybody can drop off when they need to. We will absolutely in about four minutes, no matter what, to get ready for the next session. I've got a couple of quick questions for you, Jessie and first of all, fantastic comprehensive proposal. I got lots of questions about, can I get this deck?
They're going to ask you all kinds of questions, because you've covered an entire deployment from end to end, and there's a million questions along the way to specific ones, your interaction with caregivers, and the question of health records, mm, because there's a in the United States, it would be called HIPAA. I don't know what it's called in Canada. But, can you talk to how do the caregivers who are notoriously resistant to change? How are they dealing with this and what do you deal with the health records?
Sure. Yeah, great question. and so to date, we've focused almost solely on corporate.
So we're still in kind of our working through we've, we've accomplished our first year of automations and all of our work for the most part has been centered on the corporate functions. So we haven't actually dealt with Weave.
Interacted with the medical systems but from a financial or corporate function perspective.
So we haven't we haven't really gotten into those into those areas yet. And I look forward to exploring that in the in the in the second year.
Sure, Yeah, well, what about your thoughts on, if you had to write 0 to 10 fear of automation across the system right now, where do you think? I know it varies depending on who's, but how would you rate that?
I would say, probably kind of in the 40% range. It's insignificant, and it needs to be thought through. And I think the concern isn't around automating the work that people don't like doing, they love it. Right? The quality reliability and the impact on their jobs.
And I think you do need to speak to each of those, very fair point.
I think you could even read a book on that, anyway.
So, one other question about interacting with the health records, did you, do you touch, Do you need the actual, you know, security health record stuff?
We have not. Now, I don't understand. We've gone through the privacy impact assessments for all of our different animations.
And that, in all cases, we can interact with those just like every other enterprise system interacts with health health records, but, you know, in the context of we're automating the work that people are doing today.
The work that they're doing today has already gone through those processes, and I've been reviewed. So, no, I don't, I don't foresee any, any issues at all.
OK, I want to thank William Fuller, who's online for asking, he's just Ask a series of great questions.
It's like, I wish I had ask you to pay them to do that, but I did, and he's just asked some great questions. Jesse, thank you very much for taking us through that. Healthcare is obviously important on a personal level. It's massively important, considering how big the industry is across the entire world. You've given a great example of how to use that, I happen to know, personally the Blue Prism Rock Solid Company produces fantastic products and their documentation is incredible. And, no, I'm not getting paid for that. I just worked with Blue Prism several times. You know, who said that in Europe? So thank you very much. Jesse, for everything you have, if anyone wants to reach out to Jesse is contact information, is on the screen, which I'm about telling to press stop sharing.
So, look at it quickly, and Jessie, you'll probably get the phone rang, and people will connect to you on LinkedIn.
Sounds great. Look forward to it. Thank you. Thanks. Everybody won't be back in 10 minutes, top of the hour for the next session.
And that's it.
Program Director, IT Analytics, A.I, and Automation COE,
Alberta Health Services.
Search for anything
February 21-23, 2023
March 21-23, 2023
December 8, 2022
11:00 AM - 12:00 PM ET
December 12, 2022
11:00 AM - 12:00 PM ET
January 10, 2023
10:00 AM - 11:00 AM EST | 03:00 PM - 04:00 PM GMT
January 12, 2023
3:00 PM - 4:00 PM ET
January 19, 2023
11:00 AM - 12:00 PM ET
February 2, 2023
3:00 PM - 4:00 PM ET
February 21, 2023
3:00 PM - 4:00 PM ET
Watch On-Demand Recording - Access all sessions from progressive thought leaders free of charge from our industry leading virtual conferences.Watch On-Demand Recordings For Free
Courtesy of Nintex Pty's Paul Hsu, below is a transcript of his speaking session on 'Improve employee productivity during and post-COVID by ...
Read this article about HP, Best Achievement in Operational Excellence to deliver Digital Transformation, selected by the independent judging panel, ...
Read this article about BMO Financial Group, one of our finalists, in the category Best Achievement in Operational Excellence to deliver Digital ...
Read this article about Cisco, one of our finalists, in the category Best Achievement of Operational Excellence in Internet, Education, Media & ...