Eliminating Manual Transcript Processing at the University of Cincinnati

We are thrilled to have Molly with us today from University of Cincinnati as she shares her story and how they've implemented data automation and been able to eliminate some of the challenges related to manual transcript processing. And my colleague, Julie, will lead that conversation and that q and a with Molly as she shares the story of the University of Cincinnati. So a few things before we begin. We'll do some introductions. Just quickly, my name is Matt Sternberg. I'm the director of customer advocacy.

I'll be moderating a little bit and also checking the the q and a, making sure we're getting as many questions as we can answered. But we wanna start with the conversation with the University of Cincinnati and with Molly, and we'll talk about how they've implemented data automation, why they did it, and the results that they're seeing. Then we'll get into the live demo. I think that's always an exciting part, seeing is believing, and what does this actually look like? How would this actually look like at my institution? And you'll get to see it in action, And then we'll reserve some time for q and a. Please don't hesitate to ask your questions throughout.

You might have a question for Molly or for us. We're gonna reserve some time at the end. We'll probably take a few questions throughout. We'll try to get to as many as we possibly can, but use that q and a box, which leads us to our housekeeping items. Your phone lines and computer audio are muted.

You know, you can't come off mute even. You can't ask us questions. So the way that you can do that is through the q and a. So use the chat to introduce yourself, but the q and a, that's where we want your questions. That allows us to track them really easily.

We've answered this one. We haven't answered this one, so use that q and a box when you do have a question. And I always say this. If you have a question and you're thinking, should I ask that question? Please do. Someone else on the webinar will be glad that you asked the question because they're probably wondering the same thing.

And that's part of the beauty of this webinar is we get all these people in the room. Oh, I hadn't thought of that, but that's a good question. We're all together, and we're all gonna learn from each other today. So please do ask your questions. We'll get to as many as we possibly can.

But with that, I've done enough talking. Julie, I'll pass it over to you and Justin to do some introductions, and then we'll get to know Molly. Great. Thanks so much, Matt. So great to have everybody here, especially you, Molly.

We really appreciate your time. Let's go ahead and go ahead and do some introductions. So, Molly, I'll I'll let you go ahead and go first. Sure. Hi.

I'm Molly McTownan. I'm the assistant vice provost and university registrar here at the University of Cincinnati. And I'm Julie Wakefield. I'm the manager here at. .

. In, on the Pathway sales team. I re. . .

I'm originally from Pittsburgh, PA. I saw some some in the chat, from Pittsburgh, but I'm based out of Columbus, Ohio. I've been here for about ten years now, and, I've been with Parchment for four years. And then alongside me is Justin Raynor. Go ahead, Justin.

Introduce yourself. Yeah. Thanks, Julie. My name is Justin Raynor. I'm the senior solutions engineer here at Parchment responsible for all things technical and how we integrate our products with systems that you have on campus.

I'm here to show it off to you today as well. Thanks, Julie. Yeah. And as Matt chatted a little bit already and talked about, really, the goal today is to hear from Molly and the University of Cincinnati story, to really give you more insight on the challenges that they were facing specifically with manual hand entry of data from transcripts. And, Molly, I was just kind of thinking back over the last several years and how we work together in conjunction with your admissions team, specifically first on looking to automate transcripts from Parchment Received to Slate.

And then fast forward, I guess, probably about maybe a little over a year ago, you had shared with us another challenge that you are facing. Can you talk us through maybe more specifically what you are up against with processing transcripts? Sure. Absolutely. We kinda had a perfect storm when it came to transcript processing. The credit evaluation team moved under the registrar's office in October of last year.

And at that point, we decided we needed to really evaluate efficiencies. We looked at a previous tool, and it was determined that that wasn't going to meet the needs of our university. And then a big component was the retirement of the state articulation and transfer clearinghouse, also known as the ATC. The ATC was a requirement for the state of Ohio, and we needed to send XML transcripts to Ohio public, and our SIS systems were set up to automate and intake those XML transcripts. The state of Ohio sunsetted our ATC, which then means every Ohio public XML transcript we received went manual.

So that had to be very unsettling for you. Very unsettling. Quite a large volume of transcripts that were automated, were now manual, and so that went on to our admission processing team to take that load of manual entry. And so we knew we were not sustainable in that in that manner, and we needed to find a solution. So just stepping back a little bit, you.

. . If I'm understanding correctly, you were used to already having data automatically delivered at least from the public institutions in Ohio and then automatically coming into your SIS. That's it. Okay.

So Yeah. You know, that that obviously was a glaring concern for you all when having that go away. And then, you know, was there a percentage of PDF transcripts that you all. . .

Like, was that. . . Did that come into play as well that you were looking to, you know, potentially find solutions for and and automating that process? Yeah. The whole thing is that we needed to automate.

When our processing time is sitting at thirty, forty five, sixty minutes per transcript, and we receive on an average every year, We knew, and we were at, I think at one point in time, a sixty day backlog. We were not meeting the needs of our students. And so it was just one of those breaking points that it was time to step back, evaluate, and find a different a different path for us. And I'm just just out of curiosity when you say not meeting the needs of students, what what type of impact does that have on maybe your role specifically as a university? You know, what what does that look like? The student impact is significant. When you're behind forty five days to sixty days of transfer processing.

. . And some things to keep in mind as well, it's not just our transfer students, it's also our first year students. We have in Ohio, it's called College Credit Plus, CCP, but nationally, it's known as dual admission. We have a high volume of those students coming in, and their transcripts were caught up in this process as well, meaning we had students registering for classes that they didn't need.

We had advisors who were unable to properly advise our students because we didn't have evaluations completed. We had students who were placed and were on math sequences because we didn't have the credit from the other dual enrollment schools. And it really brought down our students' trust in us as an institution because we weren't able to give them their evaluations in a timely manner. So it sounded like it was kind of like a downward spiral effect. Not only just getting that information in, but the effect that it had, like, on the students and the.

. . Their perception even of the university. And you're trying to get these students to, you know, to to come or they've already committed potentially, and you're you're looking to put your your best foot forward, if you will. Am I understanding that correctly? That's exactly correct.

We want them to have the best experience when they come into the University of Cincinnati. We wanna let them know what they've been evaluated for, what credit they've received, and to make sure our advisors are also set up for success in those advising appointments so we can map out the student's journey from start to finish so we can get them to degree. So it's a starting block, and it was a big one that we were we were stumbling on. Gotcha. So really needed a a solution.

And that that kind of, I guess, takes us to, you know, the next piece of this. You had shared with us that you had looked at some other OCR options, you know, knowing that ATC was looming and then just even that backlog that that you had already. Yeah. I should say ATC retirement was looming. So Yeah.

You had mentioned you looked at other OCR options. I believe maybe even tried to implement another tool unsuccessfully. So I'm just curious, you know you know you know, I'm sure it's probably scary to start looking at other services as well having that, you know, background. But I'm just curious ultimately why you had decided on the Parchment SmartPanda Raptor technology, and then, you know, what was that decision making based on? Absolutely. We heard about Raptor from other institutions, so we started to do our homework.

We're an Oracle PeopleSoft Campus Solutions 9. 2 institution, and so we went out and talked to other institutions who are PeopleSoft who implemented Raptor. We wanted to know what implementation was like, what results they were seeing, and honestly, lessons learned. And out of those conversations and exploration, we were like, this sounds like it's a viable tool for us, because other institutions were saying it was a game changer for them. Great.

That's great to hear. So it sounds like, yeah, just having that experience Yep. Of, you know, talking with other references and institutions who have you know, gone forward with it really seemed to make a difference for you all. Am I understanding that right? Absolutely. It absolutely made a difference for us.

Hearing from other registrars in particular on on the success they saw really put it in the forefront for us of an optimal solution. And so that's when we went down the path of, okay. Let's get a demo. Let's see what it looks like. Let's talk about what implementation would look like here at UC.

And so that's how it all started. Great. So that, I guess, really leads me to the next piece of this and and your workflow and, you know, what that looks like specifically. Certainly, there's a lot of different ways that you can move documents into an OCR technology, but this is how UC has constructed it. So I.

. . I'd love to hear a little bit more about this process and and why you decided to do it this way. Sure. So we needed something that was out of the box.

We needed something that already had implementations with Slate, where Slate Institution and PeopleSoft, and we needed something that didn't need modification. We also needed it in a quick and efficient manner, as you saw that we were up against that backlog that we were experiencing. And so how we are set up at UC is everything starts through SLATE, it implements through SLATE, and then it goes to our Raptor, and then it feeds into our SIS. For us, it's a partnership with our admissions office as well as our registrar office credit evaluation team. And so this flow works wonderfully for us.

It has been a roaring success in what we are seeing with our CRM to Raptor to PeopleSoft. Great. And I and I. . .

I'm sure that, you know, starting a new project like this was probably exciting, but also at the same time, a little bit scary, maybe. So I'd love to hear scary. Yes. Is that is that fair to say? We were we were doing this during, like, heading into peak. So we were starting this project sixty days out from our peak.

And so, yeah, it was scary because we knew we were going to have our flood of transcripts coming in in that May, June, July time frame, and that's exactly the time frame we were implementing. But it was. . . I have to say, I say this to anybody who asks us, it was the best third party implementation we've ever had, period.

Not just on the student record side, but period at the University of Cincinnati. And we were given a sixty day window of implementation, and we know that that always isn't. . . You're not always able to make that window.

You're not always gonna hit those days. We hit the date exactly, and it was actually the University of Cincinnati who paused on putting it into production because we still were adjusting business process. So implementation for us was just smooth, absolutely fantastic. Wow. That's great to hear.

That's that's amazing. So a little bit above and beyond what you expected then, is it? Yes. Yes. We we we weren't sure what to expect, and it blew our doors off. We still talk about the support that we received from the implementation team, and that implementation team is still in contact with us.

So we any questions that come up, us questioning if we're using the tool to the best of its ability, we always get those questions answered. And within maybe twenty four hours or less, it's it's great. That's amazing. Yeah. So let's talk a little bit about the results, specifically of what you're seeing, if you're if you're able to share any of that with us, and really, you know, what the impact of this tool, how that's affected your team and you and those prospective students.

So if you could walk us through that a little bit, that would be great. Absolutely. So we kind of phased approach implementing Raptor. We've pulled in transcripts a little bit of a time. Now we're we're full in.

Our transcript processing time is below three minutes. . . Four minutes. At this point, we're at, like, three point five minutes per transcript.

Really. . . You've really timed this and studied. And I think you had mentioned maybe you started, like, just with, like, the dual enrollment transcripts starting out or those CCP students.

So if you could walk us through that a little bit, that would be great. We started with transcripts that we were familiar with. So a lot of the Ohio public transcripts that we knew, we started first because if anything went weird, we knew it looked weird. We started with our CCP students as well. And then as we were comfortable, because you have to get acclimated to a new tool, the team has to get acclimated to something new.

And so as they became comfortable, we were able to expand and increase how we were evaluating the transcripts and they were going into Raptor. So that helped. Each month, our transcript processing times have dropped. So they continue to drop. Our processing time continues to drop.

We're now down to about one to two business days for transcripts to come in and go out, which is absolutely amazing. I think another key component is even though we turned this on during peak and even though we were still getting comfortable with the tool, we were able to clear a backlog, and the backlog was about sixty days within ten days. Wow. Yeah. That's really mind blowing if you look at the numbers on the screen taking, you know, that that processing time from sixty minutes per transcript.

I know we have four minutes up here, but you're saying just even under that amount of time. And then, you know, really exceeding that goal of the transfer evaluation of five days and now one to two business days to get Yeah. You know, that that completed. From a student's perspective, I. .

. I'm sure. . . I don't know if you ever hear when things are going really well.

Right? It's usually whenever they they. . . They're screaming likely whenever, you know, maybe things aren't going as well. But I'm just curious, like, you know, from the university perspective and and and just, you know, potentially, I guess, your internal processes, do you see that, you know, as being more efficient? How does how does that all look? Oh, a 100%.

I always use our advising community as a marker of how well we are doing in this space, And our advising community is just singing the praises of how we were able to get our transcripts in and processed, evaluated, and out to our students, and the time frame in which we're able to do it. Because they were our most vocal during those periods of time when we were in the backlog, because they were meeting with students and they were hearing from the students of where's my evaluation, what am I getting credit for, how is this going apply to my degree, how long do I have the degree. And so since we've implemented, those questions have really, really, really dropped to almost just nothing because we're able to get things out, because we're able to. . .

Students are able to see what we've got. Our advising community for us is our big marker because of their interaction with our student. And so it was. . .

Its impact is still being felt. It's still one of those, wow. I can't believe we were able to do this, and this is where we are starting from June to to January. Wow. That is great to hear.

Really appreciate you sharing your story. I know you're sticking around throughout the call here. We're gonna transition just to give a little bit more background on the tool, and then Justin will come in, for the demo. So, thanks again, Molly. But wanted to just share just a little bit more background information on the technology.

Through the partnership between Parchment and SmartPanda, we're really able to, as Molly described, provide a powerful solution for transcript automation. You know, I know she talked about this providing a streamlined process that really allows you to reallocate resources, give clearer data to students, not having that laborious task essentially of a human being hand typing in information that, you know, quite honestly, may be prone to errors. As Molly shared, the process is faster. You can get information in students' hands sooner for better overall student experience. And I know Molly really discussed the workflow at University of Cincinnati.

Really, that data automation raptor tool can fit anywhere in your workflow. It could process both high school and college transcripts, post data to different internal systems like, you know, Molly had mentioned PeopleSoft. You know, we work with many banner colleagues, schools as well. We post data back to CRMs, to Slate, and also to Degree Audit tools as well. So, really, there's a lot of flexibility there.

And then at this point, I'm gonna stop sharing here. I'll turn it over to you, Justin, so y'all can, see a live demo of the tool. And while we're transitioning over to Justin, I think it would be good to tackle a few questions as well because we've got some phenomenal questions in the q and a. So wanna highlight it's system agnostic, so a number of different questions. What CRM, what SIS Can work with virtually everything.

So I think that's a good thing. If you don't use PeopleSoft, that's okay. If you use Banner, that's great. So wanna highlight that that it is agnostic and can work with a lot of different workflows as well. So we have two primary purposes.

Right? Some people use this for transfer. Other people use it for undergraduate admission. Some people use it just for admits. Some people use it for all transcripts. But, Molly, wanted to direct that back to you.

Can you clarify, is it students that are admitted have made a decision, or is it students that have applied? Where. . . What exactly. .

. What transcripts are you evaluating? Great question. Right now, it is for students who have been accepted to UC and confirmed. But because of our success, we're reevaluating that and plan to probably in the near future open that up. Great.

And then someone asked the implementation time frame. I think you answered that already, which was about sixty days, and it was pretty much right on target there. And then in terms of hours, someone asked about man hours. Like, how many hours do you think it took? And I think at the spirit of the question is we got a lot of projects. We got a lot of things we wanna do.

Sixty days is one thing, but how many people do I need to to make this happen? Can you give us a frame of reference for the amount of effort it took within those sixty days? Absolutely. It was a small team for us at UC. It was less than seven people. It comprised of folks from the registrar's office, so two from the registrar's office, two or three people from our admission team, and then our student information system, business systems analyst, and they were about two, and a developer, so two on that side. So.

. . And I can't say this was a 100% of our time. It was just a portion of all the other projects that we had going on at the time. So I can't say it was a huge manpower lift on our end.

Great. And then there was one. . . Oh, what information are you actually loading into PeopleSoft? I think that's kind of an interesting question for people.

What data is actually going in? The course equivalencies for us are going into PeopleSoft. We're pulling in course, the grade, credit hours. Things like that are going into our external ed tables and the equivalencies. Alright. Well, I think we've got a number of questions we'll try to get to at the end.

Keep the questions coming. Justin, it's all yours. Thank you very much, Matt and Molly. Thank you so much for giving us that insight. I know another question that I did see pop up in the chat was, you know, non parchment transcripts as well.

This tool will take any transcript, and I even emphasize, like, official or unofficial. Like, if you're doing a review process up in front, you know, you could throw that transcript in and get a a temporary evaluation versus maybe an official one as you see here on the left side. The goal is just to get documents into the workflow. And then that's where the CRM or document management system is usually handling those documents to our data automation tool just to help funnel everything in in one smooth process. And that, looks like this with our SmartPanda partner here and their tool, Raptor.

This is the name of the AI that they give this application, and the screen that I'm on right now is actually your ability to drag and drop files, from your computer into the software if you so need to. We just talked about, you might have the funnel of documents flowing from your CRM or document imaging system, but you can also just take any transcript you need and drop it in the tool and process it. And here, the AI will evaluate and compare this transcript against other templates, layouts, blueprints of how it's seen other transcripts and the data that it's pulled off in previous instances to try and give you the most accurate read possible. This will just take a second or two to finish up. The other piece while that works is our dashboard.

This is where we funnel in those documents from sources. We often work with the CRM slate as you see here, but, really, we can set up a Dropbox or SFTP, and you can even funnel those documents directly from your parchment receive premium account into this dashboard. The goal is to collect these documents so they're already processed and run by the AI. That's what these numbers are, is the courses that have been extracted. And I can show you that.

If we go back to my transcript here, it's already, done processing, and we can tell by the colors that have been added to that black and white document. These show the courses that have been identified by that AI in green. And the other text that has been captured has been highlighted in red just to show the user, here's other things I found, but they're not coursework. We wanna draw the user's eye to that other data. But we still may utilize this data even though it's not coursework, be in the name of the student or a social or other markers that help us, determine what's valuable on that document.

Here you can see, ATP. . . Oh, I didn't mean to right click there. ATP, one ninety eight, practicum one, and the hours and the grade have all been pulled off here on that first line item on the spreadsheet.

And as I highlight and move down these course titles, you can see that highlight move and tell the user where did this data come from. I need to double check because I'm normally typing all of this information into the next system, and now the AI has produced it here for us. We also can click into the spreadsheet, and this will draw an arrow over to wherever that data came from. You can edit any of this information if you find it not to be accurate or per your specifications. This allows you to click in, and, we could change this to practicum one if we wanted, the name of that to be different or if you saw other text on there, like honors or something.

You can add whatever you see is prevalent to you. You can delete coursework. If I don't want principles of athletics, let's go ahead and delete that row. It's not a course we're gonna evaluate today for that student. And if you, were worried that the AI missed any courses, you can add new rows before and after and operate this like your standard Excel workbook or Google Sheet.

You can drag and drop to auto fill features. I like to do a little control c and then control v on my keyboard for copy and paste. And my favorite is to actually grab things off of the document. With this lasso tool, you can draw a box and capture lots of text. If you need, like, a whole course title, like basic of composition, we can highlight those three words and then click to paste it on over.

You can also click on individual words like wellness and then click, and it'll paste that as well. So, hopefully, operating the dataset that has been produced is really easy for users to make sure it matches what they wanna move over and what would normally be typed over, and take minutes upon maybe even hours, for. . . Especially when we see, this transcript with all of this green data.

And once we have all of the data in our satisfactory format and we wanna send it over to PeopleSoft, Banner, Colleague, we work with lots of different SISs, the main matching components to make sure all this data lands in the right spot is the external education ID of the organization that you store that other institution in your system. So it's often a C or a FICE code, and then we have the student ID for that individual for it to land on. Those can be typed in like I just did, but we also have many members as I'll, quickly pull up that workflow again, here, where they're collecting those documents in a system and storing that document on a student's application or profile. And that now means that that data is stored with that transcript and can often be transmitted with the document when it's batched over to us. So not only do we get the transcript file, but we get important elements, like an external organization ID or a student ID that will auto populate in here when we get those documents from those other systems.

Just further increasing, less you have to type in or or manually check when we can get those from that system. Other times, we connect to the SIS in different APIs or connections that allow us to search for these elements based on what we found on the document. Like I mentioned, the name or the social might help us get a student ID. Now that all depends on your SIS or connection. We'd love to talk about that with you in in terms of how that workflow best flows between those different systems.

And you saw me, edit this spreadsheet and take out some coursework that maybe you didn't wanna evaluate. We can also prime the system to do that for you. We have a lot of our members who choose to ignore, certain data elements because they don't need to review grades that are a d, f, or a w. We can help you with these rules across any of the different data elements you saw in the spreadsheet, as well as we capture other elements I will get to in just a moment. But to show a good example of these grades being ignored by the AI, I have a example from Carroll Community College here that I utilized to show, this student took a couple of courses that we're not going to, take in.

This personal finance course is already in the color red because the AI has, determined that that grade meets the rule. We don't need a d, so we're gonna just leave that off the spreadsheet. It's not over here. On the right, you'll see we move from intro to computer information systems to environmental science and then down to meteorology. And this, make sure that you get that double check from the AI that it's grabbing what you intend to review.

We don't need, this w, this withdrawal from sociology here, but we do want all of these courses. Those all have good grades. And then there's this introduction to business we're gonna skip as well. I use this example because it also has transfer credit on it, which is another thing we commonly ignore for most of our members. Some of them choose to capture this, which we'll represent in the spreadsheet and identify its transfer.

But the AI has been trained on not just finding the right text, but understanding that text and seeing that institution transfer. That's a string of words that it now knows the following information is probably not, from this native institution. I see this t grade. I see the name of another institution, Howard Community College. These are all examples on how the AI is trying to understand what it sees in front of it so that it can pull the right information over to the next system.

And now on this transcript, the student transfers back. So we have this c grade that we're gonna go ahead and capture as well. I'm gonna move over to my dashboard where we have the, incoming documents from those other systems. They're already prepared here, and they've been, grabbed by the AI with all that coursework. They show the student's name and the school that it came from as well and even some statistics over here on the right.

These are how the AI grades itself and gives you an idea of what you might be opening up into. The confidence on the far right is how well it understands that this is a transcript. There are a lot of OCR tools out there, and it's not a new technology. That's optical character recognition OCR. And, those systems are programmed to see what's on a document and try and pull out good information, but very management intensive.

For this tool, you don't have to set up any templates. All we need to do is plug in the transcripts and plug out the data into the next system. But it's very targeted on that document type transcript. And so this confidence is saying, hey. I I know that this is a transcript.

It's got grades and GPAs and, sections and sessions that I can understand. The filled cells, the next column, is how well it populated the data, how much of that transcript, was extracted, and are there any blanks that we should be concerned about? So this 91% is a high school transcript. And in my demo environment, I have my columns still set up with the catalog here as, we saw on my college level transcript. And the high school doesn't really apply a lot of catalog fields. So that's where you see these blanks coming in, but that is 9% of the spreadsheet missing to give us In a lot of our live implementations, we will customize these columns so that they accurately represent what data you want to extract.

But in my example here, it helps me show where blanks might occur depending on how you set up that spreadsheet. The other thing, with this high school transcript is we have some other fields that I've, set up to capture. We call those raptor fields. Anything that isn't coursework that you might be concerned about in the evaluation process, we know you're still typing in off of the transcript, things like the class size, the rank, the GPA. Maybe if a degree was earned, a start and end date for that student, a graduation date as well are all common ones that, members customize with us so that we can help pull the right data in addition to coursework and fill in all the things you'd be typing into the SIS today.

Also, with this high school transcript, we have a lot of different grades represented. I'll give our colors back here by just processing this file. It's a little bit older, but, we can still see that this first course is an English one course that has two grades, a 93 and a 99. It took place over the two semesters here, and the student got one credit for this. In our spreadsheet over here, to represent this data accurately, the AI has split this into two different line items to represent both terms, both grades, and divide the credits up by half here because we see a one, so they each get point five.

We do this a lot with quarters and trimesters as well, point three three, point two five. It's all about understanding how the document is laid out and giving the data in a workable format to the next system. Then we further customize this based on what members need. If you're looking to target certain subject areas or rigors like AP or honors, those are all examples of things we can add as additional columns so that you can make sure you identify important coursework or do GPA recalculations that might be happening in that process today. So members will work through this dashboard in order to review the documents that need extra attention that are coming from those different sources.

So this acts as a central place for multiple users to come in and review the documents they need to by type and work through, open them up, and simply push that data to the next system. By working through those documents one by one here, they're able to quickly move them over and then get answer students, to students faster. What we then work with schools to often do is automate that process further. After utilizing the AI, getting comfortable with it, understanding how well it's doing at grabbing the data that you want, you might wanna set benchmarks or automation pass for these documents. Because it might be redundant to every.

. . If, you get a lot of Plano West High School transcripts and you open that transcript up, everything looks good, and we just hit the button. We can run reports and statistics to show you how often that activity is happening. We can do it by school or by these percentages and give you an idea of how much effort you need to review in those documents.

And if you say, hey. A 100% has always been good for us or, you know, we've seen 10 Plano West High School transcripts in the last week and we haven't touched any of them, we've just pushed them on, then we'll prime the system that based on those metrics, it just fires it off to the SIS or back to your CRM wherever the best destination is. That automation means you have less documents to review. We're gonna send those automatically to success, and you would find that document information in the SIS and be able to work on it, evaluate transfer credit, get that answer back to the student faster. That's usually the main goal of most of our members is, you know, I.

. . I'm gonna work until I trust the AI and see that it's doing what I expect. And a lot of our members are seeing that statistical gain, through this dashboard, and we have some stats on that that I'll bring up here. Across our members, we've run over a million documents through, this test, application report.

And, the average processing time, for a lot of those documents that needed manual review were eighty five seconds. And that's, as Molly mentioned, you know, sometimes there's other pieces to evaluate in the transcript. This speaks specifically to time in the Raptor application reviewing any transcripts that they need to on that dashboard. And when they do open that up, 95% of the time, it's two or fewer changes that are being made to the data. Both of these are important, numbers for us to track to make sure our users are getting great benefit and, expectations are being met, there but are also ways that we train the AI on the edits that are being made within that spreadsheet.

So we know the next time you get a transcript from that institution or other members that are using this tool receive receive that similar transcript. They know that it's read and understand and and better knows that, transcript and the layout so it can accurately get that data better. That's what we call a crowdsourced AI, meaning it's learning from how everybody's using the tool and getting better across the board. That brings me to the end of my presentation, as to the time savings we might expect, within that Raptor application we call data automation. You'll see a lot of those names at Parchment, and through SmartPanda here.

But we've loved our partnership, with this company in order to help provide that missing link that we've had with many of our members. We've been great at delivering secure official documents. Now we wanna help you utilize those documents much better. So I look to hand it back to Matt, Julie, and Molly as we continue this discussion. But thank you so much for your time, everybody.

And I know we've probably have some questions that we can go through as well, Matt. We've got some phenomenal questions. I think we've got some quick hitters that I just wanna make sure people get the answer. We likely covered a few of these, but I wanna double click on them. Can the system tell the difference between a college and a high school transcript? Yes.

That is something that if you load it in manually, you would want to check the box or not. But, typically, when we get it from those other systems, we can determine based on what's coming in, how we set up the integration. So we sometimes use other data elements. But if you'd. .

. We're doing like I did in the beginning and drug a file in manually, that's where we have this, checkbox. Oops. So if I did that and we have this one, this turns on the high school AI versus leaving it off as a totally separate AI, which is for the college transcripts, and we want them to be targeted towards those different document types. Another question, does it calculate credit conversion from semester to quarter and vice versa? That all depends usually on the system we're handing it to and how you're inputting that data today.

Sometimes that's done by the system we're handing it to. That conversion is already something that's there, and we're playing off of the existing, rules that you have in play. For other members where that's a little more challenging, that usually happens after pushing the data. So we would set up rules that say, hey. When I take in.

. . I'll pull up that high school transcript again. When I take in these, semester grades, I only. .

. I want one credit. So we'll do conversion of data. That is a customization that we do per member, so we really wanna know how you do it at that institution. But the, user will still validate that this is represented accurately over here.

And then when we push, we look to transform it so that we apply all the rules correctly. But that is a common thing that we've done for members. And, a few people have asked this question, and and, Justin, I. . .

You can maybe answer it based on your experience with schools and how they do this. And then, Molly, if you have anything to add, I'd I'd love to hear it too. But, really, what people are asking about is if we don't have something in the transfer rule database or is. . .

There's a course we haven't reviewed, how do we handle that? What's actually the process? So when we say, like, transcript processing, we don't have a built in equivalency or rule for this. How do we do basically the manual review there when we don't have the equivalents or courses in the system already? Molly, if you wanted to talk about, how you might think about that process for sure, I'm happy to, talk about some other members that we've seen as well. Yeah? Yeah. If you wanna go ahead. Sure.

So our job with this tool is usually to replace the manual effort of getting that data into the next system. And then if equivalencies are there, we can, often look into those tables or try and match up that data and shortcut work. But there's still the role of the individual administrator or reviewer of that transcript in the SIS or next system to go and say, okay. Here is what was automatically taken from that transcript and input, and now I need to make a decision on it. And if there's no equivalency there, then I might take the time that I now saved from having type all of that information over to search for an equivalency or find, that course in my system and make a tie so that the next time that that transcript comes in, that tie is there and that work has already been saved.

So for a lot of our members, it's actually creating the time to do that equivalency and make those matchups. But the handling is still that process that exists and acceptance. I think a lot of our members would love to get to somewhere in the realm of auto admit, you know, that courses come in, everything matches up, and I just immediately fire off. But this is one step in the process. Assigning equivalencies is another, that, you know, helps towards that goal.

So if they don't exist, just wanna help you make those, but that's done outside the system today. We do have some work with members to help make some or suggest, different equivalencies. That's a little bit as we're trying to grow this AI utilization, but this one's targeted on text extraction and priming you to make those decisions. Molly, if you wanna add anything, I welcome that or Julie or Matt, but, that's where I see the conversation going today. We're the exact same way we handle it.

The credit evaluation team handles the equivalency, but because we've been able to save time, we look for UC equivalency. If it's for UC, we're structured slightly different than if it's a 1,000, to handle it within our office. If it's not, it goes out to the college. And so for us, it kinda goes just in a hold. We do the equivalency and then go in, and we're at about less than four to five days for that if it's internal to the office.

We've got another question on international transcripts. Give us the skinny on international transcripts. We get this question every single time. Yep. It's a great question.

Everybody's wondering about it. Yeah. I mentioned high school, college, very common transcript types that we see here in The United States, but military as well as international are ones that we hear often come up in this conversation. We've seen a lot of those different document types. And, you know, we're really good at college and high school transcripts.

You're gonna see maybe really strong statistics when thinking about documents we've seen a ton of. When it comes to military or international, these numbers for just, trying to make a tie here might be 80 or 70% in terms. . . There's gonna be blanks or it may not follow the same format.

But a lot of the right text has been extracted. We might just do some more cut and pasting. We might move, pieces around to make sure it's accurate. But a lot of that text is usually still available in that spreadsheet, for a user to manipulate, move around, and get what they need to. But they're usually translated international documents.

We get those usually from WES or other services that do that, candidly, from my engineering team. That's because I don't have any foreign transcripts if, you know, that's something we just need more database info on and give the AI the chance to read those other languages. But today, it's it's primarily been in English. But we do see those document types all the time and run them through to help save time on even trickier ones like that. This also takes scans, pictures of transcripts.

It's not gonna be as great. It'll do just as good as your human eye will in trying to grab some of that stuff. But it's still, you know, something you can throw in to utilize. And that's really our rule. If it's a transcript, we'd love you to try it in the system.

Yeah. It's a definite great starting point to use this tool. And I know seeing some of those transcripts, they are hard to decipher just, you know, from a regular human eye. So having the tool at least get you the right components, and then, certainly, there'll be a little bit more adjustments on those, foreign transcripts, you know, aside from the domestic ones. We've got a few questions.

How much does this typically cost? Great question. It's the golden question. It it depends on the size of your institution and the implementation process. So we. .

. The cool thing about this tool is it might look slightly different for every institution depending on your use case. Right? Are you using it for undergraduate admission? Are you using it for transfer? What's the workflow of how we integrate with you? So that's something that we would have to take on a case by case basis. And if you are interested, we'd love to have a conversation about what your intended use for the data automation raptor tool is, we'd be able to give you a much more accurate quote based on this. But that's something that we can talk about during the discovery process.

Would certainly love to talk with all of you in more detail about that. And then we we had another question. Does this just handle PDF? There are actually features that allow you to consume data files as well. So if you're in a state that has, you know, a common transcript initiative or you're sending EDI or XML, that's certainly a feature that could be incorporated as well in taking the data file or a PDF. That is a feature that can be activated as well.

I'm just highlighting on the screen too, Matt. Png, JPEG, TIFF, and as I mentioned, a screenshot, you know, image of those can even be uploaded. So definitely different file types, but I I will reiterate what Matt said too. We work with. .

. You know, when you get data files, from systems as well and you need to map that into the appropriate system, that can be tricky sometimes, dealing with the native piece. And so we do do that with members. It's a little bit different part of this tool, but would love to, you know, talk about those document types in our conversations. And just to add that that would be an add on part of the solution, not specifically what Justin is showing today.

We've got another question from Misty who's wondering about can multiple people process in Raptor? Does it have to be routed through one person? Then she says, I see the save for later button if my browser crashes while I lose all the updates and editing that I did if I didn't click the save for later button. And I think at the spirit of these questions are, you know, I wanna make sure if we have multiple people using it, that they're not stepping over each other, grabbing the same files, and all that stuff. Can you kinda walk us through. . .

And maybe, Molly, I would love to hear, is it one person dedicated? How do you, actually divvy up the work? Yeah. I think from our perspective here, we wanna create a good dashboard that multiple users, can access, and you see that I've locked this transcript right now, so that nobody else would be on top. And that is an important feature that we want one good read and process to go over. I did see in the chat earlier about duplicates. We do look for that in our integration to see if the coursework already exist when we go to deliver it, but we want a user to get a good read send it over.

Right now, my other colleagues could not access the documents, that I'm on in the system here. But it also, I know, like, we talked about maybe this. . . Somebody saw this save button.

This is actually to export the information that you have right now. It's a little bit of a different logo, than your traditional, like, save. But the edits that you're making in this dashboard of documents will be saved if I, was to get rid of this, you know, and and we'll change this to four point o, for example, if that's what we were doing, save. Oof. It's it's there.

And then if I close out of it and bring that one back up, then we'll see that change is is still applied. You can also see now my colleague, Julie, on the call, has opened up a document so that we can see I can no longer get into that one. So it's designed for multiple users to be in here and to find and filter the right documents that they need, maybe in a certain status. I talked a lot about waiting for review, under review, or moving it to success, but we also might use an exception queue or a processing queue, to help move it to a different team member if they've got a review. We just want the workflow to match what you all do internally to help the right eye see the document and then push the data on, keeping in mind that that data can still be worked with in a system like PeopleSoft or Banner or Colleague once we move it there as well too.

This is really that review process to say, did the AI type what all I would type? I'll pause with that answer and see if Molly wants to add anything or Julie if there's any, from other members you've heard too. Absolutely. We're we're very similar to what Justin described. We have at least seven people at any given point in time working on our transcripts within Raptor. I think another fun feature that we enjoy as well is we can see user and user time on how long it takes to get through as well.

So that's another component that you see find valuable, but we're on best practice similar to Justin. We'll take one more question here. And Molly, I wanna pass it over to you, and then we'll wrap up today's conversation. I do wanna be sensitive to time. This is being recorded, and we will be sending out the recording to the people that registered for the webinar.

See a few people saying, oh, I had to jump on a little bit later than I wanted, or I want my supervisor to watch this. So we will send that out. But someone asked kind of a great question. We covered a little bit of this earlier, but what other vendors did you look at with OCR slash AI technology with similar features? Why Parchment? Because I think that is. .

. I don't wanna hide behind. . . Like, there are other solutions out there, Molly.

What stood out about Parchment and our partnership with SmartPanda, and and what do you think differentiates us, both in terms of product, people, implementation, all the rest? We did. I think that's a really good point is we did look at other vendors. We even tried to implement with another vendor, and the labor left was significant without the return on investment from our opinion. We have a long standing partnership with Parchment since 2016. We heard the value of Raptor for for about a year, almost eighteen months.

We saw its ease of integration. It it checked all the right boxes for us at UC on. . . The integration was already there.

It's out of the box. The support is there. The trust in the partnership with Parchment is there. It just was a comprehensive product that stood above the rest in our view. And we really appreciate you sharing your story.

Like, as someone that works with our customers a lot, like, when you say it, it means a lot more than when we say it. Like, we think we're so great. You should work with us. But to hear your perspective is valuable, especially in education. We know that referrals and references, you know, the experience of your peers and colleagues is extremely valuable, so we really appreciate it.

We try to incorporate that as much as we possibly can in these conversations. And so in the chat, I put a link to book an additional conversation with us to learn more about your institution, what's driving you, why you wanna do this. So book a demo with us by clicking that link in the chat. Unfortunately, we couldn't get to every single question in the q and a, but we'll try to follow-up with all of you individually. And this is a slide that I wanna highlight too because product is one thing.

Right? You know, does this have all the features that I want? But it is also about the people. Right? If if you just could see a demo, does this have have everything I want? That's one question. But people matter. Customer service matters. And I wanna share this slide because as someone who cares about customer satisfaction, our average rating when we survey people is 8.

5 out of 10. We're really proud of that. or higher in terms of whether or not they would recommend us to a friend or colleague. satisfied with the people they work with at Parchment. renew their contracts with us.

So I highlight all of that as just we are proud of the way that we work with institutions. We are not perfect, but we try to get it right. And I think I only highlight that as, hey. If this. .

. If you like the the features that you saw today, we work really hard to make sure that it becomes a reality for you that you actually see the benefits that we implement on time and that we're there to serve you throughout the process. It's not just an install. See you later. We'll talk to you when your renewal comes up.

We really work hard to have a continuous relationship with everybody that we work with. And then as I mentioned, if you wanna continue the conversation, I put the link in the chat. Or if you're already working with your customer success manager or an account executive at at Parchment already, please continue that conversation. And then a shameless plug for our podcast. If you've never listened to the Credentials Unscripted podcast, we'd love for you to check it out.

We don't talk about our products and services. It's mostly thought leadership around student pathways and credential innovation. We just had one on competency based education. We have one coming out on how you can do credit for prior learning for short term and noncredit credentials, which is something I know the registrar community cares about. We have Melanie Booth from the Higher Learning Commission on for our next episode and Ian Rourke from Pima Community College talking about CPL for noncredit credentials, which I think is really interesting stuff.

I am the host of the podcast, which is why it's a shameless plug. Check it out. And if you've listened to the podcast, let us know in the chat. If you listened to the podcast and didn't enjoy it, please do not let us know in the chat. Alright.

I joke. We're having fun. But really do appreciate Molly joining us. Join me in the chat thanking Molly for sharing her experience. And, again, we'll follow-up with all of you and try to get to as many questions as we as we can, and we'll send out the recording as well.

But thank you for taking an hour out of your busy day to join us, and we look forward to continue the conversation related to data automation. Anything I missed from the Parchment team or from anybody on the call? Thank you. Thanks, Molly. Thanks, Matt. Alright.

Have a phenomenal afternoon. Thank you, Molly, and we'll talk to everybody real soon. Goodbye. Bye.