Tell us a bit about the initial implementation, such as its department or process. How long ago was that?
Sure, so I would say that this was before my time. I've been with the company for almost six years. Our previous imaging platform was Oracle, which we had contracted with ImageSource to manage for us. When that solution approached the end of life, we did an RFP and selected the ImageSource ILINX application to be the Oracle successor. That was about 15 years ago. We did implement, over 12 years ago, the initial ILINX solution.
Volume has grown exponentially since our initial implementation. Still, we receive anywhere from 20 to 30,000 actual hard pieces of mail per day, and these are these pieces of mail could be anything from a question to some people wanting to take action on their credit file, such as a dispute, [etc.]. All that information is scanned into the platform and worked through ILINX in work queues in one of our multiple contact centers, but primarily in one. We have between 3 and 500 agents that work in the platform daily as their primary responsibility.
Since that time, we have upgraded the solution a couple of times. We've put it on a new architecture. Then, fast forward, we've also expanded the platform. We've added more scanning capabilities to handle additional volumes with the pandemic. We quickly had to transition all our workforce to remote, working in three different countries. So that was somewhat challenging. That work from home experience has become successful where we've seen productivity increase and absenteeism decrease. So we decided about three-quarters of a year into the pandemic that we would have some permanent workforce remain at home in this workforce space.
The information that we deal with is highly sensitive and secure—moving to the work-from-home environment introduced risk because all of our facilities are HSA, which is high security, where no cell phones or anything like that is allowed on the floor. That can't be controlled in a remote work environment. I was asked to develop solution options to improve the security for a permanent work-from-home solution.
So [in] partnering with ImageSource, we decided to do three separate projects. The first project expanded the platform into a high availability solution that spanned multiple data centers in the experience environment. So that was one because we determined that we couldn't have a significant outage on this platform due to compliance and implications with the work stored in the application. That is currently in flight.
The second thing that we wanted was to automate many of the repeatable transactions. We are using the ImageSource Optical Character Recognition technology (OCR) as part of the platform, and the first phase we implemented was automated sorting. All the scanned images pass through OCR, and we're sorting those into separate work queues depending on the work type. Typically, we have about 35 people manually sorting these by reading the actual mail piece and then putting it in a work queue. In implementing the automated sort, we've eliminated those 35 headcounts and moved them into other work areas because we've obtained over an 80% confidence factor in the OCR implementation.
The next level that we are doing right now is to automate those transactions end-to-end rather than just sorting. So, for example, if we have somebody who says I want a copy of my credit report in a mail piece, OCR will identify that. It will process that transaction into our Salesforce CRM system and do that with no agent interaction whatsoever.
We're automating those transactions again for any of the high-volume transactions. This could be “I want a copy of my credit report,” “I want a fraud alert,” etc. Then, for things that can't fully be automated, we're automating what we can. For example, if a customer has a dispute request. We will identify that as a dispute, send the transaction information that is recognizable and known on the dispute mail piece, and prepopulate that in our Salesforce CRM platform. There we're hopeful that 80% of those dispute requests only require a cursory review of the request, so the agent would only have to submit it rather than re-entering all that dispute data.
We’re able to redact a lot of sensitive PII data. So, for example, things like Social Security numbers will be stored in the Salesforce database but won't be visible to the agent processing the transaction. Therefore, if we have a bad actor taking a video of their screen at home or having over-the-shoulder lookers in their home, that information won't be visible. That significantly reduces our risk as well as improves automation.
The last component that we have not yet started but is scheduled to start later this year is again partnering with Image Source. We are implementing the ILINX Data Loss Prevention solution. With DLP, the OCR technology views the scanned image and redacts sensitive information that the individual may have written or typed on an actual mail piece. That way, if a remote agent needs to view the actual mail piece from Salesforce, they will not have any sensitive information available to them. They can validate the transaction, but all the OCR technology-identified information would automatically be loaded to Salesforce. That would be the behind-the-scenes to process the transaction.
For those transactions that OCR cannot identify, they'll go into a reject queue. That reject queue would be worked in an environment in an office. Again, we expect that number to be less than 20% of the total volume based upon our current automated sorting success rate. So when you think about those extensions, which are pretty significant, and that we were already doing work within ILINX, that factored into decision making to extend rather than adopt new systems.
Pre-pandemic, we discussed expanding the OCR technology into automation for these transactions we currently have in flight and integrating to Salesforce. There were other benefits because we didn't have to pivot between ILINX and Salesforce to complete the transaction.
The challenge was that introducing all this automation had some cost-benefit, but it wasn't super significant because the primary workforce that works these mail pieces is in Santiago, Chile. The cost of labor there is low. Expensive automation has a much longer recovery time to recognize the benefits. The pandemic and the productivity increases realized with work-from-home gave us a solid case to make this investment.
Would it have taken much more time and analysis to pull the trigger on remote work without the pandemic?
Yeah, I had several meetings with the CEO, and this shift occurred not just with our contact center workforce but also globally. In the last two years in our history, it's been the best financial growth in the company history. During this time, the workforce shined, which caused a shift in the executive leadership’s thinking.
I've been on calls with the CEO, presenting these options, and his response to me is: I no longer have any interest in investing in any facilities. Our future is a remote workforce. We've proven this experiment due to the pandemic; we found that we were more productive. We can find the best talent regardless of where they live in the world and not require an in-office presence. For us, pre-pandemic, one of our business unit presidents used to walk the floor every Friday and count how many people were in the office. So, it shifted the company's philosophy overall in all divisions.
We started to hire remote wherever possible and continue to do so. Again, just in the contact center space, that was the trigger because this automation was not cheap– about $4 million to implement this automation. The CEO said this is a worthwhile investment based on the productivity gains. We've also expanded the contact center workforce into a virtual center this year in Oklahoma. We did training in a hotel and then sent them on their merry way to work from home. Again, it goes to where we can find the best resources.
Would you think it's fair to say that, particularly from a C-level perspective, there's been a shift in appreciation from pure cost justification to one that's more shared with value creation and taking advantage of opportunities long-term?
That president I talked about lived in Southern California, where our corporate headquarters is. He flat out said you guys all proved me wrong. We don't need to be in an office, and he has relocated to Florida.
That's interesting to hear because I've talked with other companies like this. Many of them say it's the first time that I've had a full appreciation by my executive suite that ultimately, we're in an information business, you know, no matter what business we're in. We're in the business of managing information assets, and that's a portable business, and we have to start acting like that.
You know, we wouldn't have chosen the remote work, but now that we've had that, made all those changes and investments, it's a question of we did this under some degree of duress, but now what can we do with it? How do we optimize?
Right exactly. The millions of dollars that a global company spends on facilities and just saying, OK, 90% of our workforce doesn't need a facility. That’s a significant impact on the bottom line. We opened our corporate office a couple of months ago to fully vaccinated employees, and that campus is designed to hold 5000 people. The most that we've had in the actual office has been 80.
Yeah, just on a personal level. I keep wondering long-term what will happen in the commercial real estate business. It’s just kind of remarkable how maybe it'll drive down the cost of residential real estate at some point and in cities so that people can afford to live there. It's just such a huge change.
Yeah, it's driving the cost of personal real estate up because people are moving to where they want to work now that they're not tied to an office.
So as you did all this, what were the primary aspects of ILINX that allowed you to move so quickly?
The IT was more of a risk acceptance because we're heavily regulated and have a substantial direct-to-consumer business. If we weren't able to service them, the company would take a huge hit. We genuinely want to change the perception of credit bureaus in the marketplace. So we've developed new products at no cost to a consumer, where we allow customers to self-report like their utility data on their credit report to improve their credit score. We were in the process of rebranding that, and so it was critical at the C-level that we'd be able to continue on that journey even with the pandemic. ILINX was already in use, so it was more about a risk acceptance at the C-level to allow all our agents to work from home. ILINX, in its current implementation, alleviated that risk. We were able to transition all of our workforces in the North America operations, which is three countries and globally, in my responsibility, nine countries. All in a matter of fewer than three weeks.
Was there anything other than duress that convinced the people to overcome that risk aversion?
I would say duress, as well as our brand. We wanted to prove to our clients and consumers that we were there for them in this pandemic. I wasn't given a three-week timeline. It was to figure out how we could do this and put together a plan. I presented it to the CEO, and he said, let's get on it.
Yeah, that's a little different than the ROI calculation that often exists around these projects.
Think about it. All of our workforce in our contact centers, none of them had laptops. They had zero client workstations to connect to a VDI environment in the office. So, I had to scramble, and every other company was in the same position to find laptops. You're in nine countries and need to figure out how to get these laptops to these agents. It was a challenge; I had to buy refurbished equipment. I was buying whoever had any piece of inventory at all to get these out here and then looking at in-country buys, so I didn't have to go through customs. But we were able to pull it off in less than three weeks.
So has this whole experience resulted in you thinking about whether there are some applications that you can sunset now or that you're thinking about differently than you once did?
With ILINX automation and Salesforce integration, we plan to reduce the footprint of ILINX because all the work queues would be moved into Salesforce. We're changing how we're using the platform to sort and an integration engine into Salesforce itself. Then ILINX application will only be used by the remaining 20% of the workforce still in the office to manage the reject queue. The processing logic will reside in the ILINX application; the queues and the accessibility of the data will be managed in Salesforce.
What do you know now that you wish you had known a year ago?
It wasn't anything to do with imaging. It had to do with remote connectivity. Everybody is connecting to the data centers via a VPN connection that they log onto with their laptop and then access the application suites to do their job in a VDI environment. So specifically in our countries, such as Costa Rica and Chile, they don't have the hardened Internet infrastructure that North America has, for example. Even our UK operations don't have as well-established infrastructure as North America.
Another part of this project has been how do we improve? Since this is our permanent work stance, how do we improve that remote connectivity specifically by implementing a different access method from a cloud-based VPN provider using the Google backbone to traverse the traffic to our data center?
So, the agent in Chile, for example. Instead of connecting to the Internet from their home to the data center over the open Internet, they will connect to the in-country Google Pop in Santiago. Then we'll ride the Google backbone, a hardened MPLS network, back to the data center. It reduces the Internet footprint from the agent home to the data centers. That was significant learning for me, looking at the possibility of agent connectivity in these other countries. Then the other thing that was a little eye-opening is what I'm calling the bad actors. People took advantage of having a laptop in their home, and they were doing things with it they shouldn't have been doing. We had to terminate several people just because they were abusing the privilege of having a computer in their home.
Right, that's interesting that they probably were using their phones for Internet connectivity before this.
Exactly, and many of them don't even have phones.
Yeah, the issue you talk about with both the first mile and the last mile is so interesting about how people have had to adjust. So, is there anything, in particular, you want to share about having a really good, valid, robust point solution and being asked to extend that functionality into areas that maybe we didn't think about initially?
The only thing I'll add is the C-level leaders of the experience organization embrace this. We're open to change, and we're there to look at productivity. The contact center was one of the first teams we transitioned to work from home because they're the face of our company. They needed to continue from a consumer servicing and a regulatory perspective. The CFE will fine us if we don't respond to a dispute request within 72 hours. So, they were listed as the number one resource.
We had to continue moving forward, and our executives always put the employee's health first, so there was no mandate. But then, when we were successful with the contact center, we started looking globally with ILINX. Can this be successful? And so that was also a good lesson learned from this activity. Our financial performance proved that this was more successful than an in-office workforce requirement, which was a good learning experience.