AP Automation Case Study at Nuvance Health

Nuvance Health (formerly Western Connecticut Health Network) completed an Accounts Payable Automation project integrating Kofax TotalAgility with Infor Accounts Payable Invoice Processing (APIA) and Perceptive Content (ImageNow) for increased straight through processing of invoices and less manual data entry. As an outcome, Nuvance Health saw average Invoice Processing days decrease from 45 to just 5 days.

Download PDF
Transcript

Julia Robinson:
AP alpha sort by processor. Good afternoon, everybody. Welcome to Webinar Wednesdays with RPI Consultants. Today, we’re here with Miles McIvor, and we’re going to be reviewing our AP Automation Project at Nuvance Health.

To introduce myself, my name is Julia Robinson. I am a senior project manager with RPI. I’ve been a project manager for about 10 years, and I’ve been here with RPI for about the last two years or so. One fun fact about me is I am the master of karaoke.

Miles McIvor:
I was asking around the office earlier, and that is under debate, currently. So, I haven’t seen it myself, but I’ll take your word for it.

Julia Robinson:
There are some videos floating around.

Miles McIvor:
Okay.

Julia Robinson:
My agent is removing those, but-

Miles McIvor:
Okay. Good. Yeah.

Julia Robinson:
But it’s a true statement.

Miles McIvor:
Okay. All right. My name is Miles McIvor. I’m the manager of accounting systems at Nuvance Health System. I come from a professional consulting background, doing implementations, kind of like this actually, roughly 10 years in the accounting and finance space, implementations and ERP. And a fact about me is I referee rugby. I’ve actually refereed some international stuff.

Julia Robinson:
Very cool.

Miles McIvor:
I think it’s interesting.

Julia Robinson:
It’s very interesting.

Miles McIvor:
Some other people might not.

Julia Robinson:
Very interesting. All right. So, a couple things we’re going to cover today, we’re going to review Nuvance Health as a company itself. We’re going to review our challenges that we presented at the beginning of this project, what our solution for the AP solution was, and then also review use some outcomes, some metrics, and what our results from the project was and some results from Nuvance itself.

Miles McIvor:
Great. So, this is about my company, Nuvance Health. We were actually two health systems merged into one, Western Connecticut Health Network and Health Quest. And so Western Connecticut, by virtue of its name, Western Connecticut, and then Health Quest is in the Hudson Valley, just over the border of New York State.

Currently, we’re about 2,600 doctors, 12,000 employees there. We have Infor as our ERP. We have it for HR, supply chain, payroll, and we have Cerner as our EMR. So that’s just a bit of a background into our major technologies we have.

We use Perceptive for all our imaging, as our imaging repository for all our digital images. Specifically, our accounts payable department, we currently employ eight staff and we process 190,000 invoices annually across the network. So pretty high volume for us. Yeah. It always shocks me when I see that number, but there’s a lot of…you’ve got to go into the AP team. They work hard.

Julia Robinson:
Yes, they do.

Miles McIvor:
So, this all started with…actually, we needed to assess where we were coming together as an organization, and we needed to just kind of see where we were at, right? So, we had RPI come in and do an assessment.

Everything on the screen here, they went through all of our subsystems and basically did a procure-to-pay assessment, start to finish. So, everything from the beginning, purchasing materials, all the way through to payment. That was very eye-opening for us.

We’re a hospital, right? So, we don’t do this often, right? We don’t go through our systems. But, specifically, one of the major things that came out of that was the AP department and the fact that we were manually processing everything. It’s something we’ve been doing for since the beginning of computers, typing it in, which just seems obvious. But we just didn’t know how far behind we really were and the fact that there were imaging options out there.

Julia Robinson:
Miles before.

Miles McIvor:
Yeah, this was me before. This was a year ago. I aged terribly since then. But so, yeah. So, when we look at…so we said to RPI, “Okay, we’ve got this problem. We’ve got this AP problem here, this manual process. What can we do to digitize us and get us into the future?”

This is actually me. I’m wearing the same shirt. Before we did the implementation. And as you can see, this is just one of our cubes with archive invoice images. What we do is, we’d get images in the mail. We would key them, put them in a box. I’m sure people watching are very familiar with this. And then, not in the picture is we’ve got filing cabinets as far as the eye can see with our current fiscal year. And quickly need to look something up.

So, this is what we’re dealing with. It just seems crazy now that we’re automated, but that’s what we were doing a year ago. It was taking roughly 30, 45 days for an invoice to come in and get processed. We had our stacks. Like AP department have all the images… invoices, excuse me.

We had no visibility into how long they were sitting there or where things were in the workflow. Anyone who’s worked in AP, the black hole. People call up, “Where’s my invoice? Where is it in the process?”

Julia Robinson:
Well, you did the best you could.

Miles McIvor:
Yeah. Sometimes we didn’t know. You can look at a filing cabinet, you can look in the stack. But, obviously, a huge problem when you’re talking about paying bills and keeping things current.

Any approvals or escalations for invoices were all done over email. Again, we got tons of email a day, people missed emails, didn’t read them, and just a lot of things lost like that. So, there’s a lot of churn related to that also.

We had the old AP alpha sort via processor. So certain AP processes had their letters that they processed. It gets the job done eventually, just very inefficient.

So, this is kind of like the high-level view of what came out of that assessment and what we were dealing with. And obviously these metrics, 45 days paid invoice, we have most terms of 30 days at the longest. That’s a big issue.

Julia Robinson:
So, whenever we review the challenge, so we knew that we wanted to automate the AP and invoice processing. Some of the things that Nuvance was kind of looking into was cost savings, being able to reallocate their resources to do other things besides just manually keying in those invoices. And then also the transparency that we kind of hit on just a little bit ago. “Where are the invoices? What is their status?”

So, what do we want? We want everything. And so how do we match up the technology to what the solution needed to be? So that’s where, between the assessment and the design and everything, we were moving forward.

Miles McIvor:
Right. And us as a client was concerned. We didn’t really know what we wanted. We just wanted a digital kind of end-to-end process. So, here’s the objectives actually on the screen here.

This is kind of what we had as our requirements. So, we vetted different vendors. ARCA is one of them, obviously, that we ended up going with. These were our main things we wanted. So, we wanted digital. We didn’t want paper anymore. We want to get rid of the paper. We wanted digital images coming in via email that would be ingested into the system.

If someone emailed something…excuse me, paper mailed something in, we’d scan it in the same way, it would ingest the same way. We wanted consistency in documentation across all AP departments. So, we didn’t want each AP processor doing their own thinking, keying a certain way. Wanted consistency across.

Julia Robinson:
Also, that brings up a good point with onboarding. If you have a consistent process across all for all processors, you bring somebody new on, so much less timeframe to get them up to speed and actually processing those invoices in those time periods you have set.

Miles McIvor:
Absolutely, yeah. And then we’ll talk about it later about the documentation that you say. We use it for onboarding now. It’s very helpful.

I already talked about the reduction of paper. We try to eliminate it completely, but we still have people mailing in invoices, which is fine. But we tell them, “Hey, if you email them now, we can process a lot quicker. Anyway, we’ll talk about that later.

These were our business objectives beforehand. We’re very excited to get to the end product here. We wanted straight-through processing. So, if something was a PO invoice, that’s pre-approved. Everyone signed off on it. PO invoice comes in, it’s getting straight through and paid. There isn’t delays. Obviously, the manual process was very error prone. Even identifying those errors is difficult. We wanted a system where we could see where invoices were at every step of the way, and if there were errors, be able to triage them quickly.

With any AP process or implementation, you want to go in with a clean Vendor Master. So, one of our objectives was to have whichever consulting company we worked with coming in and cleanse that Vendor Master before we got started.

So that’s basically the six things up there, straightforward, but I think anyone, looking at one of these projects, seeing these would be very close to what we had here. Yeah. This is kind of it.

Again, just touching on our Infor version, we’re version 10 in the cloud. That’s our existing technology, what we implemented on. There’s our accounts payable department. We’re at eight processors and a mail clerk that sorts the mail. And kind of the idea was we could implement automation and have our AP staff do more value-added things.

Here’s a list of a couple of things we have them doing now. We can research things a lot quicker. We can do our reconciliation and kind of cleanliness activities a lot quicker. So that was kind of our mindset. If we can improve this process and add efficiencies, moving forward we can repurpose. So, no one’s losing a job or anything crazy like that. You get to tell that. But we didn’t want to. We just wanted to improve efficiencies, and we were really able to do that, which we’ll talk about shortly here.

So, these are the requirements, what we wanted. This is just recapping what I just spoke about. We wanted imaging. We wanted OCR. So, we wanted to be able to scan in. We didn’t want a person reading that invoice and validating. We just wanted it to go through OCR to read that.

Advanced data capture. That means basically extracting information off that invoice and being able to intelligently assign it, pull the invoice number, pull the date, look for duplicates. Things like that.

An approval workflow. The email, it’s just so antiquated and just doesn’t really work. We wanted something within the solution to be able to send out, send back, route around approvals to managers, directors, so and so, based on approval thresholds. Things like that.

Reporting. Everyone wants reporting. Give me more reporting. we-

Julia Robinson:
There’s never enough reporting.

Miles McIvor:
There really isn’t. I mean, we joke about it. We have the canned Infor reports for AP, which gives you a certain kind of point in time, but nothing advanced. I want us to know, like how many are we processing, how days aged are we, what’s our payables coming up? Like how much do we have outstanding? So that was one of our big requirements.

Training documentation, huge. We touched on it already. And the Vendor Master Cleanup, absolute necessity for a project like this.

Julia Robinson:
Perfect. So, we went through RFP process with Nuvance and we proposed a solution. We’re going to review in these next couple of slides coming up the technical solution that came from the assessment that we did with Nuvance. Kofax TotalAgility for the invoice capture and data extraction, data validation, which was a big part of this. And then also Infor’s APIA solution to actually process and get those invoices sent out for approvals.

Miles McIvor:
Yeah. And I’ve just got to note here, this here is the RPI plug coming up right now, but when we did the RFP process, a lot of pushback I got from my senior management all the way up to the CIO is we usually have a tendency to hire these huge companies and they come in and do what they do. And there was an apprehension with a smaller company, relatively speaking, RPI, 100, 150 people, to go with that.

But the level of comfort I had with it, smaller shop, that the consultants were extremely knowledgeable in the business. They’re experienced. And we just talked about yours and the consultants we had on this project. Very knowledgeable. You get that kind of…you feel like a real customer, not just the number.

So that worked out really well, and I can’t wait to kind of, yeah. It’s nice to be top on those things.

Julia Robinson:
Thank you.

Miles McIvor:
Yeah.

Julia Robinson:
Perfect. All right. So, a high-level overview of the very simplistic solution here. So, we have the Kofax KTA solution, which invoice capture. So, it’s grabbing that invoice. It’s extracting the data. The users are validating that extracted data against the invoice image.

Yoga Connect, which is a proprietary software that RPI developed, it transfers that data over to APIA, where that invoice is then processed and sent for payment. So, it’s Kofax over to APIA, every step of the way you’re able to view that invoice and see how long it’s taking in each queue or each person hands.

Miles McIvor:
Yeah, it’s really cool. It’s kind of like a live lookup. So, if you’re in Kofax here and you need to pull information like a PO, for example, you type in the PO number on the invoice, it’ll populate all that information directly from Infor. So, I thought that was kind of cool.

Julia Robinson:
Yeah. Very cool. So, our Kofax TotalAgility part of the project, by our spectacular Brian Ayres and Richard, our KTA professionals here.

So emailed and scanned documents, we were able to capture those. We have an AP Invoices at the nuvancehealth.com. People can bring those in or send those in that way.

We were also able to OCR paper statements via scanning. So, the scanner process didn’t necessarily go away or…I mean, we built on the scanner process to be able to process those things, so they weren’t manually entering them. OCR in the email. PO, non-PO invoices, statements, and credit memos.

The data extraction was by document type. So, every type of PO, non-PO statements, credits, they all have different data, part of them. So, dates, names, vendor numbers, line items, all that stuff was able to be extracted off of the different document types.

Miles McIvor:
Yeah. And the Kofax solution actually is able to identify which it is by the data on the form, which is really nice.

Julia Robinson:
Pretty smart.

Miles McIvor:
Yeah. Very cool.

Julia Robinson:
Pretty smart. So, additionally, the data validation against Infor, so as Miles was just talking about, it’s a live look-in to your backend system. You can type it right in, and you’re able to say, “Oh, yes. This. I want to go look up this PO.” Well, you have all that data in real time.

Miles McIvor:
Yeah. Very helpful.

Julia Robinson:
Automated approval routing. So, we talked a little bit about the alpha previous approval routing, but they really wanted to get to a point where anybody can work any time. And we wanted… That’s how we get some of links of time down. and we’ll talk about that in just a second.

But also, buyer queues and vendor maintenance queues. So, setting up special queues for those people before it was held through email and now it’s all within the same system.

Miles McIvor:
Yeah. Anyone from an IP department watching this will know, if there is not enough money on the PO, the PO is closed, you send an email. Like just craziness. So, the buyer queue enables us to flag it, send it to the buyer, it goes back to a buyer queue, so the buyer can work that invoice instead of ATP. So again, we eliminate the email component.

Julia Robinson:
So, on our APIA solution, which was implemented by Sean LaBonte and Jeremy Stoltzfus, something really cool that we did as a part of this project was, we customized the HR structure that was built on the in-port GHR. It was routing and maintaining that tables for users. So, I’ll kind of let Miles…he gets really excited about this.

Miles McIvor:
Really excited.

Julia Robinson:
Yeah.

Miles McIvor:
It gets more exciting.

Julia Robinson:
Yeah. More exciting. All right.

Miles McIvor:
So basically, APIA does…Infor has its own out-of-the-box solution. It’s a table, right?

Julia Robinson:
Yeah.

Miles McIvor:
You put in the person’s name, their manager. You maintain this table. We just…I said previously, we have 12,000 employees. I’m not maintaining that table, so how do we do that? It’s kind of…this is a big part of the project, actually.

So, what we did, we think, “Well, HR does this. Why don’t we have them use their structure, which they maintain, and use that for our approval routing.” So, what it does, it goes back, looks at GHR, looks in the person’s manager or director if it’s over a certain threshold, and does it automatically. We don’t have to update a table. It updates the entire table on a nightly basis. Pulls in any changes, personnel changes, and we don’t have to maintain it. So that was really cool. And like you said, this was a custom solution. This wasn’t something that out-of-the-box. Like RPI actually built this for us.

Julia Robinson:
Right. Yeah.

Miles McIvor:
So very cool.

Julia Robinson:
Yeah. Definitely able to customize a solution to whatever the client’s needs are. So that was a really big one for us, especially for this project.

Processing of invoices, credit, memos, and utilities via the APIA system. We did build some exception processes for manual resolution. One that I can think of off the top of my head is the duplicate queue. If a vendor kept emailing the same invoice over and over and over again, the system was smart enough to pick up and say, “Hey, this is already either in a queue working or it’s already been resolved.” And so, there were these exception queues that we put in place for them.

Miles McIvor:
Yeah, just a note on that. What we found after implementing is actually about 20% of our invoices were duplicates, which is…I showed you the numbers earlier, 190,000. You can imagine 20% of that being duplicates. Someone will email something, and they’ll mail it in. They don’t get paid that week. They’ll email it again. This takes care of all that. So, we’re not getting these duplicates stacking up. So that duplicate kind of identification thing is a huge time-saver.

Julia Robinson:
Perfect. Also, the interface with the Infor ERP, this is the payment posting part of it. So, it’s a live look-in. So, we directly interface with that and we can…it’s agnostics. So, whatever your backend system. In this case, we have the Infor side of it. But we have that live look-in on the payment posting. And then, additionally, manager notifications. Before, there really wasn’t any of that. Now, if Miles gets sent an invoice, he has to approve it, and he says, “I’ll do it tomorrow,” and he doesn’t, he’s going to get a notification of the payment. “You need to go take care of this.”

Miles McIvor:
Exactly. And it’s really simple. If you’re a manager and you’re getting these approvals, there’s a link in there and you click it, you can see the invoice image. You can put in your coding, and that’s basically it. It goes in for payment. Just from what we had to this, it’s really cool. It really modernizes the system big time.

Julia Robinson:
So, I want to talk a little bit about our project plan and implementation. So, we signed the Statement of Work and the Agreement with Nuvance in March of 2019. And they had a fiscal year deadline to go live with this project by 1 October.

It’s a pretty big implementation, overall. A lot of moving parts that went in there. So, October deadline, that wall could never move. So, anything that happened…of course, our project was perfect, and we had no issues whatsoever.

Miles McIvor:
Yeah, 100%.

Julia Robinson:
One hundred percent. But one of the things, we had to condense implementation timeline because we spent so much time in design to make sure that we were getting it right the first time. That’s RPI. We always say, “Hey, we’re going to do it right the first time.” But in this case, or, well, it did happen in this case, but we had to spend a lot more time in design and a lot less time in implementation. So condensed implementation timeline, one star, do not recommend.

Miles McIvor:
One star.

Julia Robinson:
We got it done and everything was fine, but it was quite stressful for both teams.

Miles McIvor:
Yeah. Julia and I had a lot of honest conversations during that time. No, it was a hard deadline and it’s stressful when we put that in place. Like things change. Our fiscal year end was on that date and…but we-

Julia Robinson:
We had to hit it.

Miles McIvor:
And we did.

Julia Robinson:
And we did. It was perfect. It was on the button.

Miles McIvor:
Yeah. We got on the money.

Julia Robinson:
On the button. Some of the professional services that we had in scope, process analysis and solution design, this is where we spent a really big bulk of our time in the beginning.

We had all those subject matter experts in the room. We made sure that the processes that we were putting in place were best practice. We wanted to make sure that the people in there weren’t just reproducing manual processes before by utilizing a new system, just because it’s fresh and shiny. We really need to implement best practice there. And we did.

Obviously, our software implementation, custom development. We did create some custom aspects of this project. We had multiple training cycles. This is something that at the beginning of a project, because training seems so far away, we really need those multiple training cycles, multiple testing cycles.

And not only did we do onsite training a couple of times, but we also did some guided UATs. So, user acceptance testing, where we did training, and then they did hands-on, which you can show them all day, but until they get their hands on it and actually use it, they’re really not getting that training they need.

Miles McIvor:
That was actually one of our requirements. We didn’t want online. We didn’t want…we wanted a button to say…teach us this stuff. And that’s what we got. And we were…that was part of our testing. So, as we were training everyone, they were learning the system. So, you’ve got…it was a good bang for the buck with that.

Julia Robinson:
Yes, it was.

Miles McIvor:
Highly recommend.

Julia Robinson:
We also came onsite for go live. I think this helps a lot with the anxieties of a go live. Having that technical resource there that’s kind of been leading the project physically there, whenever you flip that switch and you’re in your production system and they’re doing it, they’re going to have questions. They’re going to have anxieties. They’re going to, “Hey, is it supposed to do this?” And so even though they’ve been in training for three weeks, four weeks, whatever, all of a sudden now it’s real. So, having that person there was really important to us.

Miles McIvor:
Yes. I had high anxiety, and Brian Ayres was my best friend for a couple of weeks there, the implementation lead there. So, thank God for him.

Julia Robinson:
Another thing that we took into consideration, and it was a part of the scope, is just that post go-live stabilization. Just because you go live with a solution doesn’t mean that immediately everything’s just going to be rainbows.

So, there was about six months of post go-life stabilization. And now that the system is doing good, small tweaks, small changes, retraining of employees just as they need to be reminded as to what they’re they’re doing and then what they’re processing. So that was a big part of our maintaining that relationship with them, maintaining the platform and the product.

Miles McIvor:
Yeah. Just the post go-live stabilization. We actually have…obviou