 Good morning, everyone. I figure we'll start on time because it would be wonderful if we can have some conversation, because really what we're here to do is share about what we're doing and to learn from you all. Honestly, there's really bright lights here. You can't even see people in the audience. But first, I want to thank you all for coming. When I looked at the program, and we're up against ACLS and funders, I was expecting this to be one of those things where you don't want the cameras to pan to the audience. It's on video, and we can do our presentation. But so I'm really, really grateful for how many people are here to listen, and hopefully to talk with us about what we're doing. What we're going to do today is, and one of the reasons that I'm here, is to really give you a sense of why we're doing what we're doing, how we're doing it, and hopefully to have a conversation about how we can do things better to serve the community and what we're trying to do in this program. So really try to give you a sense of the field. So some of it will almost be regressive in the sense of things that everybody knows, but that we want to just try to ground the conversation and why we're doing what we're doing at Edithica and JSTOR right now in this sort of transition as it continues from the analog world into the digital world. So first we'll go all the way back. I think most of you know, but in case you don't, I've been here since the very beginning when we first started JSTOR in 1995. So I have all these memories. And so sometimes you have to listen to some of that old stuff, but in the very beginning, the immediate goal was to convert the back files of scholarly journals into digital format that was really built. I mean, I don't know that we really thought of it, but it was really what we ended up doing was starting to build a digital infrastructure. We didn't call it that. We didn't call it putting collections into the cloud. We didn't call it the things we call it today, but it really was at the beginning of that process to look at physical collections that needed to be held locally and expensively locally, and which were difficult to use and get to. And we really digitized them to save resources for the community and in effect put those collections into the cloud so that they would be less expensive. And that what we learned in that process was that they were in addition to being shared and less expensive for the community, they were more impactful, they were more useful. And that actually expanded even far beyond what our expectations were. But really at the core of it was this notion of the infrastructure changing. The infrastructure to support the physical access to collections was moving to being digital. And we were kind of at the very beginning of that. There was no network at the time. We originally thought we were gonna digitize these all for CD-ROMs to make it smaller and make it less expensive. So the world has changed quite a bit like that. But some things still are the same. Very quickly, obviously already users don't wanna use print in some cases, they really gravitate toward the digital. But the students that are coming in the future, they're starting very young. They're using these things. I have kids who just finished college and when they were younger, they didn't have access to these things. So these things are changing dramatically. We all know that already students want different things but these students are not gonna be doing this. I mean, we know that. And so when you work in the libraries, you have to address those needs. You're already addressing them. You're already wrestling with them. And so I'm gonna kind of step through a little bit about like what do we know about what's happening in this transition? And much of this is not new. I just wanna ground us in the conversation in why we're doing what we're doing. So increasingly, so there's one set of principles that in some sense are about the content and the changing nature of the needs for different kinds of content. So increasingly, faculty wanna teach and with their and their other libraries digitize primary source collections. Like primary or they wanna get at them or they wanna use them for research. And students are being asked in their assignments. It's not enough to summarize book chapters and journal articles, make a unique argument. Show us a unique argument with unique materials. Use primary source materials. So people are required to do that. And journals and books are increasingly available electronically both when they're first published and through projects like JSTOR and other digitization initiatives. But still those primary source materials, they're increasingly becoming digitized, but obviously many, many, many, many items and objects are not yet digitized. But even those that are digitized are very difficult to find. They may be available via Google, but if you do searches, they might appear 1,000th on the list or for a variety of reasons, they're difficult for students and faculty to find and discover and use. And so in that sense, they are very similar, at least to my way of thinking and I'm biased, that they feel a lot like journals in 1995, right? They're out there, they're available, they're valuable, but they're really hard to get to. They're hard to find, they're hard to use in the process of teaching and learning. So increasingly obviously, obvious statement. Everybody loves convenience. People wanna do their work digitally, however they can. They wanna work on the network. They don't wanna have to go travel to different places to get access to collections and they wanna do their work digitally. Libraries need to provide the collections and services that they use to provide physically through the network and in digital form. And as we know, there are economies of scale for a lot of these technological capabilities that sort of pressure them up into these mega companies, whether it's Facebook or Google or whatever, and increasingly things happening in the cloud can be shared and so there's more and more pressure on that and I think that institutions have done a great job and are working toward, okay, what part of those things go as Lorcan used to say to the network level and which of those things that have to remain local and to what degree do institutions or libraries continue to focus on that local part? What's that local part that we actually add value to by being local? All that being even more challenged by what happened with the pandemic and moving us away from these sort of physical interactions to these sort of zoom-like interactions but increasingly that had been sort of one of the areas that libraries would focus on is like, what are the local things that we can do and what can we do with the network that will both save resources and take advantage of those economies of scale? I think AI will threaten these local aspects but we can talk more about that later. But effectively if you're in the library or you're at the institution working in scholarly resources or in information resources, like every strategic choice, you kind of have three choices, right? You can do something yourself, like build it, build it yourself. You can collaborate with others to work in a shared way or you can buy the capability or license the capability or do something like that from a service provider. So it's often said, build, buy or borrow or some form of that is a choice strategically that institutions need to make as they wrestle with these kinds of choices, especially in areas that are subject to economies of scale from technological connections. So what does that look like? Now this is not our slide. This is not created to go over and describe. This is a slide to illustrate the complexity of the situation. So this is a slide from a research one library. They have this on their website about the challenges that they're wrestling with across the top. You're thinking about collections. We have to store the collections. It says storage at the top. Then it says we've got to preserve the collections. It's got preservation. We need to produce these collections, digitize them, add metadata to them, et cetera. We need to manage those collections, figure out how to manage them, where to put them, and we need to provide access to those collections. Those are the columns across the top and then these sort of like silicon chip looking connections between everything, our different services or capabilities that they're trying to implement to do that work. And highlighting both the complexity of it and the need for it for institutions. So again, this is just to highlight the complexity and the different components of what a library might be thinking about doing. And I think the point that I would want to make in this simplest possible way is just that everything that a library used to do in the physical world in order to manage and provide access and make their collections really useful in the teaching and learning process has to be done in the digital world. I mean, you have to have the material. You have to have access to it. You have to put it on the equivalent of a shelf somehow. You have to manage access to that shelf, try to make it really, really helpful. You have to provide a security perimeter around it. You got to make sure people aren't stealing it. You need to make sure people are using it in the right way. You have to guard people's privacy. All of those things that needed to be done in the physical now need to be done in the virtual and the electronic. And obviously it's a transition so you have all your costs associated with the physical while you're trying to now put your resources into this new future. And so you've got to kind of do both and you're managing that. And increasingly over time, sort of the reason for the picture of the little kids with the iPads is that more and more of the work is gonna have to be digital. And I think there'll always be obviously physical collections and needs, but that will become a diminished amount in the sense that you've got to be able to do more and more on the digital side. Now thankfully a lot of that work comparatively is way less expensive, right? I mean, in terms of building a building and having the infrastructure like in the aggregate, it's less expensive, but it's still costs. And when it's an additive cost on top of your physical, it's gonna be very challenging. So academic and research libraries are in the midst of this major strategic reinvention to imagine how they support the next generation of instruction learning and scholarship. The collections need to be increasingly digital, not only what's licensed from others, but the collections that you all own and have cared for for a very long time and invested a tremendous amount in by keeping them for all that time. You've kept them for a reason. How do you then make those accessible not only to the students and scholars on your own campus, but obviously if you can, to students and scholars all over the world. So at the same time that you're doing this, there are challenges. There's, you know, your institutions have fewer resources. There's a lot of pressure. Libraries need to demonstrate their value to the home institutions in a variety of ways and be important. And, you know, the challenges of moving into the STM world and resources going toward there. How do you show and demonstrate the value, especially for humanities and humanistic social sciences? And again, libraries need to make, build or buy the systems and platforms that they need to manage that transition to basically create the digital version of what is offered by the physical library. So when we look at that, so what does it look like from where, you know, where I sit or where I stand, having been working in this area for 28 years, I look at it and I say, as I said earlier, gosh, this feels very familiar. Like these challenges are quite similar to the conversion of print journals into the electronic form. We built all this infrastructure and when I say infrastructure, I don't just mean the sort of technological infrastructure of bits and wires and things, but also things that are part of infrastructure that are really important, like the awareness of JSTOR as a thing that people go to. The use, the many, many faculty that are assigning students to go there, the impressions of that, the awareness of it. Then of course, all the bits and bytes and the infrastructure that's there. And as an offer-profit organization, we're at a point where we say, what are we now doing with that? That a lot of what we did, if you think about taking a journal and digitizing it and making it available on the web and it's searchable and a PDF is deliverable, that is now a commodity, that is now available all over the world. We basically assume that every journal article, book chapter that's in JSTOR is available somewhere on the web for free, searchable by Google. That's just the truth. And we can't, what we then say when we think that is, okay, well if we need to support this infrastructure going forward, we have to add more value than that. We have to be doing things that are valuable to users and to institutions that make it worthwhile to support that beyond just, you know, flat search of images or book chapters or whatever it is and search and delivery. So how do we do that? How do we take our value in a nonprofit that's built on that foundation to do more things? And that's how we think about what we're trying to do. So we've been building this infrastructure for over two decades that results in JSTOR on the access side, Portico on the preservation side. As we said, JSTOR is in the middle of the sort of research and learning workflow. So a lot of people are already there. It's a starting place for a lot of students and faculty. The preservation platform has, you know, almost 20 years of work and evolution through multiple migrations and so we know we can do that and preserve materials effectively. And we're then taking that infrastructure and trying to make it available to institutions. So we've built all of this. How do we make that infrastructure available so it's less costly to get more of these collections recognized and evaluated and used than they would otherwise be at the same time that we can reduce the costs of doing so. So in effect, we built this infrastructure, make it available to institutions. That's the institutional side and the value from a user side is the collections that have been cared for by your institutions for many, many years, which are really hard to find. Now are in the workflow of students that already go and use JSTOR for a variety of reasons. We think if we do effectively use machine learning, other kinds of techniques, we can make that material more accessible, more useful for the students. So you have value on both sides, just as it was with the original JSTOR. We can reduce the costs, increase the impact, and then have a fundamentally different experience for the users. So that's what we're trying to do. That's how we're going about it. Actually doing it from a practical point of view leads to a lot of challenges. And as we've been working on it, we're gonna share with you some of the things that we have learned and have happened. What are some of the things that are working and maybe not working. And then kind of open it up to talk about how can we do this better and how can we think about the way to build that infrastructure in a way that's most useful for your institution. So I'll introduce Bruce to kind of walk us through some of the practical experiences what we've learned from trying to do what I just described. Thank you, Kevin. Good morning, everybody. Yeah, so we went through a process for actually a couple of years of talking to libraries about the struggles they were having with their special primary source distinctive collections. And we had a lot of conversation internally about what could we as an organization actually bring to the table that would be unique, that would be different. Didn't wanna just build another institutional repository, there are plenty of those out there. We wanted to see if we could do something that would potentially be transformative. So we started very modestly just talking to some libraries that say, hey, we're thinking about sharing our infrastructure. Would you be interested in putting a couple of collections up on the platform? We kind of wanna see how they perform. We really had no idea in any sort of measurable way. We knew from our work in the past with books and with obviously the back files of journals that when things, and with images from art store, when things got on the platform, they got used a lot. And they got used a lot more than they typically got used previously. So we kind of went in with the idea that potentially that could happen with this type of content as well. So we started, we didn't have a business model, we were just sort of like exploring. We had a set of hypotheses that we wanted to sort of test out and this sort of notion of having a couple of institutions put things on began to grow and we ended up actually kind of formalizing a pilot project with institutions too so that we could do a more sort of formal evaluation of what we were finding out. So that led to what was many of you may have heard of before we called the Open Community Collections Initiative. That was basically the pilot project. It was sort of learn as institutions put things on the platform. And our hypotheses were really threefold. We had heard a lot from institutions that as Kevin said earlier, their content was siloed. It wasn't in the research workflow. It wasn't at the places where people actually start their research. So we had a hypothesis that putting stuff onto the JSTOR platform would actually make that content more discoverable around the world, quite honestly. So we were working with institutions primarily who had the rights to make content openly available. So the idea was to put the content on the platform as it was gonna be open access. So you didn't have to be a JSTOR participant to use it. Anyone on the open web could use it. We wanted to test that out, how people used JSTOR in the open space. A second hypothesis of ours was we'd heard from researchers for some time that even when they were able to find these primary source materials, they were, it was a disjointed process. It wasn't connected at all to the secondary literature that they wanted to reference and sometimes reference this content. And they really were interested in finding a more seamless way of making that available. So our hypothesis was that having these primary source collections on the platform along with the really great secondary literature that's already on JSTOR would actually make the research process more seamless and more efficient. And if we were able to connect those primary source collections with the secondary literature in really effective ways using machine learning and artificial intelligence, that would be really valuable to researchers. The third hypothesis was that, and we heard this very early on was that one of the big issues that libraries explained to us they were having with these collections was they did not have a long-term preservation solution. And that many of them were having stuff backed up at their institution, on servers, on other things, but that really wasn't the managed preservation they were looking for. In addition, many were very concerned about the increasing costs of actually trying to make this content usable as research techniques change. So we heard about people needing to meet accessibility mandates from their state or from their institution. There were new discovery techniques that were happening that this content just was not optimized for at all. There were texted data mining. People wanted to get their stuff for text analysis and the content wasn't really set up for that in many ways. So there was all these things happening with how research wanted to happen on this content that institutions were just saying, we don't have the time, we don't have the money and we don't have, frankly, we don't have the people anymore to actually do the work that we would need to do to make this content usable in the new environment that we're in. So our hypothesis was that obviously we have Portico that we could probably make some, have some help or provide some help on the preservation front and that with what we were doing with our own content with AI and with machine learning, we potentially could help this content kind of move forward into the current research regimen. So during the pilot, we had about a little over 300 institutions put almost 1,900 collections on the platform. So we had everything from postcards from World War II to gear books from institutions to electronic theses and dissertations to monographs. We had things like the, from CRL, the South Asia Open Archive, put their content on the platform. From the Catholic Resources Research Association, they put the Catholic News Archive on the platform. Canadiana put some of their content on there. It was just a lot of people testing different things out to see what the usage looked like, to see what our infrastructure might provide. And what we found was interesting was that obviously the usage of these collections started to grow over time, what we would expect, but also that people might have started by putting one or two collections on the platform just to test it out and they were finding it to be successful enough that they started putting more. So when we started, we'd have, I think the average number of collections coming on the platform was like two per institution. When we ended, it was like six per institution. So for some people, that's a pittance of collections at your own institution. But what it was signaling to us is that people were finding it valuable enough to try to put more things on and to learn new things. Obviously as this went on, the conversation about preservation became more pronounced and we decided that we were going to begin to offer preservation for these collections through Portico, which was very different because Portico had originally been set up to do the long-term preservation of published content from publishers. And that's a very different thing from people putting their digital collections in. They have a different use case and so we've been learning from that quite a bit about how Portico needs to change somewhat to meet some of those needs. And then we found that people were actually interested in the cataloging tools that we had built for people to build collections. So that was very interesting things that we learned just on our own stuff during the pilot. On the use side, some of the things we found, I found quite interesting. So I would say these are things I found interesting. First of all, over 60% of the use of the item requests that were coming to this content were coming from the open web. They were not coming from, we have 13,000 institutions around the world that licensed something from JSTOR. They were not coming from those IP addresses. So some of this could be inflated because of the pandemic and obviously people were working from home and may not have been authenticating through their IP address. They may have been a faculty member at an institution. But the fact was that the open part of what we were trying to do seemed to be working, that this stuff was getting out on the open web. We found that institutions from around the world were using it, so we had over 13,000 institutions from 225 countries access this content, which means it was getting the global reach that a lot of people were actually looking for. The fact that we were kind of interested with JSTOR as a starting place for a lot of people in the humanities and social sciences still prove itself out here. We saw that about 43% of the usage came from people who were actually starting at JSTOR to find that content and then we found that about 31% of that usage was coming from Google. Now that percentage is actually higher than what we see on the journals and books on JSTOR. The Google use for journals and books on JSTOR is actually in the 20 to 25% range for our books and journals. So Google was having more of an impact on this content. And one of the things that we learned during this was that we were taking the objects and the metadata. And one of the reasons that we were taking both the objects, we weren't just being a metadata form basically. And we learned a lesson a little bit from DPLA where they were finding that their content was not getting indexed as well as they'd like to in Google because they just had the metadata. So for having the objects and the metadata, Google started indexing this stuff very aggressively. And because JSTOR is one of the most used resources on the web period, the indexing in Google, even if it was being indexed at the home institution, that indexing was being amplified because of the linking into JSTOR and the connecting to JSTOR. So that amplified indexing I think was leading a little bit more to the Google use that we were seeing. The first half of 2023, we spent a lot of time starting to connect the secondary content on JSTOR with the primary source content. We still have a lot of work to do there, but we started to see some real impact happening when you connect these things and are able to show people articles about objects or show them objects that are connected to articles or book chapters that are in JSTOR. That's fertile ground for us to build out even more and we're very excited to do that. We also, through this pilot, made the decision that we're gonna start supporting audio and video on JSTOR. So we started supporting audio the first of October. Video I think we just started supporting last week. So now the institutions that have oral histories and have video and things of that can put that content on JSTOR. It'll be indexed. We're doing auto-generated transcripts for everything that comes on the platform. There's a lot of content that people wanna put on that they haven't done transcripts for. So we're at least gonna do that and same thing with text that comes on that we OCR. Another thing we found that was working was we had institutions who had built collections and other repositories obviously. Content DM, Islandora, Digital Commons, whatever happened to be, Ten, Alma Digital. And we tried to make it really easy for institutions to be able to move that content, I shouldn't say move, share that content on into the JSTOR channel so they could put things up easily. We found that that worked really, it was hard work on our end for doing the harvesting and that sort of thing, but it made it really easy for institutions to think about trying out the infrastructure we were building. And then we started to see people start to move, want to move things into portico for preservation. And in fact, some people, that was the most important thing for them to figure that piece out. So those were things that we found were working pretty well. We had a couple of, I'm just gonna give you a couple examples of folks who have made decisions to start using our infrastructure in probably more substantial ways. So the CUNY system, this is from a presentation that Kristen Hart, the University Librarian for the CUNY system gave at Charleston last month. But CUNY has decided to adopt the things that we're building for the CUNY system. So they're gonna be using our cataloging tool to build their collections. They're gonna be using JSTOR as a as sort of a primary access point for these collections and using portico as a preservation home for these collections for all the CUNY campuses. And that actually started from the ground up. We had one of the schools in the CUNY system, Queens College, that started putting content in during the pilot period. And they were finding very good success and what you can't see on the screen here, but what I'm actually showing is one of the things they were interested in they had digitized this bilingual news weekly called Greek News. And they were interested, would they be able to get out to researchers in Greece, quite frankly, that might be interested in this? And what they found in putting this stuff out, I had a similar experience with the University of Manchester with some South Asia content they had. And they found when they put the stuff into JSTOR they were starting to reach a community, they were unable to reach with their own website. So that was to them was really interesting. That was one of the things they tested. Another thing they tested was could they put some things up that would drive more funding from a donor to digitize the rest of a collection by showing the reach it was having, they were successful there. And actually some of their content, one of their collections in particular I got featured in NPR on something. And so what this started to build at Queens was a recognition in the organization, not just the library organization but Queens writ large that this had some branding and fundraising capabilities for them. So that led to this moving up to the full CUNY system about what they possibly could do to partner with a not-for-profit organization like ourselves to think about how we build this infrastructure together and make that available to the system. We had a consortium, the Sculpt Consortium out in California that was interested in building a digital cooperative. So Terry Oaks-Galloway, the executive director, she gave this presentation at Charleston as well. And one of the really interesting things that drove their interest was one of the things I talked about earlier which was when we started there were many institutions who were a little worried, would we simply be cannibalizing the usage from their own platforms on their own institutions? Like was JSTOR gonna be additive? Were we gonna be having the stuff in the JSTOR channel was it gonna reach a different audience? Was it gonna reach a broader audience? Or was it just gonna be cannibalizing an audience that already existed? And one of the things that they found in the institutions that were working with Skelk was that not only were we, was the content in our channel reaching a different audience and a broader audience, but it was actually because of the Google indexing was actually helping their, the access at their home institution from their content DM site or whatever, BPress, whatever happened to be that usage growth. So that was a really interesting output. We haven't measured that across all of that but that was one of the interesting things that the Skelk institutions found. So Skelk said, well, you know, we're similar missions. We have similar objectives. What JSTOR is building is sort of a fit for purpose approach where they can choose the pieces they want. They don't need preservation. They can simply choose to share the collections that they can choose to build them. And so we're working with Skelk to provide the services to their consortium through Skelk as services that Skelk is providing on. So we're trying to find how to partner with people in the right way. And then just another great example is Skidmore. I think I saw Marta here. So Skidmore has kind of adopted the full stack of what we're doing at this stage where they're using our JSTOR forum to build their collections. They're sharing their collections on JSTOR and they're preserving them in Portico. And we're seeing a number of other institutions kind of move into this space. So Kevin, let me turn it over to you. So we go back to this picture and just to highlight the complexity of that challenge and to sort of talk about where we are. And Joe, Lucy, I see you're here in the audience. I don't know, it's probably a year ago or something I was talking to Joe about these things and he was saying, well, we're engaged right now and across that asking exactly these questions and we're working on this. And I think we haven't done a really good job of having people be aware of how we're trying to be helpful in this area. So let me just talk a little bit about what we do that's already in this mix. Obviously, from a content point of view, we have the archive collections, we have books, we have our JSTOR images and the JSTOR digital library. We've got research reports, these open shared collections and then the revealed digital projects. So we've built a lot of content, a lot of different types. And as we said, that's moving into different format areas, audio and video in addition to images. We've created a set of tools that are operating on the platform, one of which is around cataloging JSTOR forum which was evolved out of the shared shelf initiative at art store to catalog and track art images. But increasingly that cataloging and that metadata around all content is super important and is a big part of this process. But we also have a resource called Constellate which is a platform for helping teach about text and data analytics with a special emphasis on humanities and humanistic social sciences and social sciences. You may be aware that we have a beta right now. We've released a gen AI chat, kind of a JSTOR user assistant on the platform and between ML and AI and gen AI believe that that's gonna be a really critical element in discovery of primary source materials because of the lack of ability to really make full text connections between those things and instead trying to develop semantic understanding of what those primary source materials convey or what's in them, being able to make that semantic connection into the content that we have and coming up with ways for users to be able to interrogate and question that as they would if they were talking to another human being. So really exciting things going on there in the tools area. And then in terms of our services, Bruce talked about them, preserved collections, shared collections and some of these new things that are around harvesting. The question for us right now is we have all these different pieces and we see the value of these different pieces but how do you tie them together? What's the core of the infrastructure that's required in order for institutions to going back to what I said before, you have to build, buy or manage or collaborate. You're going to be making those decisions. How does the resource, how did the infrastructure that we built fit into that? And I mentioned Joe earlier because he was working on that and he talked to me about it and it was like we hadn't gotten that message across and fundamentally I think there's a piece of this which is about the active management of these collections and materials for you and the way that you manage them over time and direct them in different things. It's different from oh, am I preserving them in a service like Portico or is it accessible in JSTOR? Like how am I actively managing these? How am I controlling these at the institution? I'm putting some words in your mouth just to test them but I feel like we haven't really got that glue in the resource that we're providing. We've got a nice collection of things here we're doing a lot of work and I'll tell you that the teams in the organization are like, whoa, slow down. We have so many things going on that we're trying to do but I think there's a core to this that is really important to building that infrastructure and actually being able to realize the benefits that are both around long-term savings and also increased impact of these materials and I don't think we've really figured that out and so our appeal is like we're looking for folks who wanna help work with us to build that infrastructure that key connective tissue to these things that we're doing so we wanted to try to share here like why are we doing it? What is the problem that we see? Do we have that right? Do you all see that problem? Are you struggling with those things or you might say no, no, we've got this solved or that's not a good way to go and that's kind of the spirit in which we're approaching both this meeting and where we're going now because at this point we have to make some decisions about how we're going to tie these things together and we need to get help to do that and we'll have to invest to do that to provide capabilities that are currently not picked up by these different tools, services and content. So I wanted to stop there and sort of open it up to you all and see if we have questions about anything. You could be around this, you could be around our Gen AI chat, whatever you might wanna ask about what we're doing right now and we look forward to the conversation we have for just 10 minutes or so. Okay, so either Bruce or I can answer questions. Anybody have thoughts or questions? People get up, you think they're coming to the microphone but they're actually leaving. That's a moment, yes. Hi, Jamie Saragosi from Stony Brook University. I'm trying to channel my special collections, the director of special collections and she tends to be very protective over her records and the record creation. And as you're talking about linking these secondary sources to the primary sources and kind of enhancing some of the metadata, I just wonder how much control the original library or the depositing library has over quality control of record creation and enhancement. Yes, so that's a great question and I think there are a couple of different dimensions to that. So I think first, the tool has to be created in a way that allows people to interact with that content directly, to see that content, to edit it, to manage it, et cetera. And I think there are probably gonna be some folks that will say we don't want you to do anything with this and that'll be something we would have to evaluate like if something could be done to make something broadly accessible, more accessible in a different way. Right now, we haven't made it so that we don't have a process right now in the way that we developed this for sort of a kind of a veto power over, like let's say you had a piece of metadata that you wanted to include, that we wanted to include that would make it more useful. We don't right now have that process in place, but that has come up in questions. So I'll give you one example where one of the institutions has built their own, they've actually created their own transcriptions of audio, and so we have an auto generation, so they want their version, and that makes sense, right? They've corrected it a certain way. So I think we're gonna have to have ways to be able to override what we're doing in order for that metadata to be controlled by the institution. What do you have? Yeah, just, I can add to that a little bit. So when institutions contribute or put content on the platform, we don't mess with the metadata at all. We map it to our metadata schema, but we don't change any metadata elements so the institution keeps control over that. What we do to make stuff more discoverable on our platform, we'll put kind of like a wrapper around that content where we've done things with machine learning and AI to make that stuff more usable on our platform, but we don't change any of the underlying metadata right now. And I think it's really important to emphasize that too that we really don't touch any of the content in these collections in the sense of like, it's uploaded by the institution and it's actually from a legal point of view, we don't touch that content. That's content that you all put on the platform. We don't change the content. When I say don't touch it, we would manage it, help preserve it, we would try to make it more accessible, but we don't actually change the content. And ultimately the idea is that the content goes on the platform and then we figure out how to make it more useful with a variety of techniques for discovery and search, et cetera. So why don't I go over here and then we'll come back. Hello. My name is Hongma from Loyola University, Chicago. I found that this is very exciting as we are looking for ways to explore a hidden special collection for research. I also wonder again, in my institution, we try to incorporate a special collection into our research and learning team for librarians to teach and research support. Could you elaborate a little bit more about how this infrastructure will fit into the general library discovery system? So I think that if we're talking about discovery layoff services, do you mean that type of discovery? Yeah, like ex-liberates, pretty much, or ABSCO EDS or any other type of infrastructure or tour. Yeah, so do you wanna go ahead and take this one? I mean, I go ahead. Okay, so right now the content that is contributed, that put on the platform, the open content, we push out to Alma, to EDS, to OCLC. So it can be, it's part of what we push out on a daily basis to the discovery providers for inclusion in their knowledge basis. Not only do people put open content on the platform, they also put content, they don't have the rights to share openly, but they can only share to their institution. So we have the ability to provide that only to that institution. We do not push that out to the discovery providers. Thank you. So we have those relationships that exist for those discovery providers also for all the licensed materials. So we already have had that relationship for setting the metadata, updating the metadata, et cetera, and so we're adding this content to that. Now, how good they are at surfacing any of that content, I cannot say. Right, we have lots of metadata issues there. So that's different. Yeah, thank you. Hi, Nadia Gossady from Washington University in St. Louis. And I am curious to what degree your users have contributed born digital content, particularly born digital content that originated in legacy formats. I don't have a specific number on that. What I do know is that about 45% of the content that has been contributed is actually image content. The born digitalness of it, I have no idea. What we're gonna start to see, because we just introduced audio and video, there's a lot, we're gonna see a lot more I think what you're talking about come on the platform that way. We just haven't seen it yet. Great, thank you. Yeah, I think our, it's probably partially just because of who we are that it's been more converted content, digitized content as opposed to born digital content. I think that with Portico, we're seeing a lot more of those kinds of challenges and sort of more web archiving kinds of questions that are coming up around how we're preserving different types of collections. But that hasn't been an area where institutions have immediately looked to us for the reasons, for kind of obvious reasons in the terms of the type of work we've done in the past, but fully anticipate if this becomes as important as we think it is that we'll have more and more of that type of content. Awesome. Well, thanks very much. We'll give everyone, Joe, do you have a question or are you leaving? I actually do have a question. Okay, great. I don't know, you just picked up your bag. Yeah, no, I was gonna go to the mic. Joe Lucia from Temple University. So here's a dilemma and I'll be honest, it's a dilemma we're still struggling with around this service and our own internal decisions, which is that our thinking about presentation and use of digital object we create is kind of segregated from our thinking about preservation. And it's been really hard to get staff involved in those two spaces, although they're sometimes the same staff, but they I think sequester their thinking. So the idea of a product offering that attempts to address both of those needs has been complicated for us to integrate into our decision making around preservation specifically. And I just throw that out there because I think part of it is needing to think holistically about the digital artifacts you produce and where they live and how they're used, but also then how they're managed through time. And we have not been able to converge on a common vision for those two sets of activities. I just make that as a comment and I think it's been a challenge for us around how to make a decision about where we wanna go with what you're doing versus what a pure preservation enterprise is offering us. No, that's really helpful. And I, sometimes internally we, so I'll say a couple of things. One is that we have chosen to have these range of services well defined as these services. And I think there is a desire in general in our experience over time that libraries need and like choice. So being able to have these services be unbundled and not be bundled is both a positive and a negative with respect to the question that you're raising. The other piece of this that's a little bit challenging, like sometimes think of these services as like the fingers and then like this core file management control is kind of the palm if you will a bit. And one of the challenges is like those things overlap also, right? So there is a part of that preservation solution which is about file management and then moving around of those files which you also have to do if you wanna provide great access which you also have to do if you wanna do some other things, security, et cetera. So that's also a hard thing to put into each of the special things and not think of as a bundle. So there's kind of both tensions going on. The desire to bring them together but also the ability to keep them separate in certain ways. So we're wrestling with that and it's that core part that combines this active management. We can say oh well we have a preservation service like Portico and the preservation service we provide for publishers is pretty passive, right? I mean we are active but from the publisher point of view they provide access to their materials and we then bring it in, we then normalize and migrate those data, we do that work but they're not actively engaged in doing that. A library with its collections wants to be actively engaged in doing that. So that's a different kind of preservation service. Now there may be really small or under-resourced libraries or whatever that would say okay we're comfortable, we trust you, we'll put our collections there and we know you're preserving it but a lot of libraries, most libraries will say I wanna have active control of that. Questions like what's happened to that metadata? How are you managing that? So that becomes a management issue that's connected to the preservation function but it's also connected to the access function, it's also connected to some of those other functions. So those become a little bit hard to parse out and develop into specific products and it's exactly that kind of engagement that we're looking for libraries to raise their hand and say we wanna work with you on this and help you figure out the right way to do this because I think we need to invest more in that file management part of the enterprise than just those services that are around the edges and so that's the moment we're in right now and we're making an appeal, we're looking for people to help with that. We obviously have a lot going on already, we're doing a lot of things that are having a lot of great impact but we see a real opportunity at a really significant level to help with the infrastructure problem. Keeping these fine people from lunch I think. Thank you so much. Thank you very much. Thanks for coming.