Afternoon Keynote: Innovation - Learning Through Personal Connection

Please welcome to the stage Instructure's chief product officer, Shirin Vijay Singhur. Good afternoon and welcome to the jungles of Borneo. How's everybody doing? Yeah. Alright. I was a little worried after a lesson an after lunch session, it's always got a little bit of a question mark on the energy level, but you all are gonna bring it. So I'm so excited you're here.

I got a chance to, you know, get to spend some time yesterday walking around, all the different sessions and get to as many as I could. And there was one that I found, that was pretty interesting. It was a session with students, and it was just a really powerful reminder of why we do what we do. Right? That's the the people that we're trying to solve this for. Every single one of us, we're solving it for learners.

So that was really exciting. And then I went a little bit later, in the night. It was a lot later in the night. How many people went to the hackathon last night? A few of you. Okay.

Cool. I went to the hackathon and the energy double was was really high there and it was kind of fun to see folks, innovating and building together with our product and engineering teams. And so as as someone who who really enjoys kind of standing up here and talking about all of the exciting innovations and all the fun stuff we're going to do. And and that's not just because I lead product at an EdTech company, but it is exciting because we're seeing a big shift, Right? We're seeing this tech enabled learning shift that is now the norm. It is becoming the standard.

And if you look at the data, right, you see that, you know, there's a bump that we saw, in the post COVID era, but that's something that's sustained. We're seeing the sustained growth. We're seeing it become the new norm. And even though we see this as k twelve data, we're seeing this pattern persist. We're seeing this pattern persist, in higher ed.

We're seeing this pattern persist in, you know, places outside of higher ed and beyond. And that's really exciting. So when I was out here last year, we spent a lot of time talking about AI. Right? I don't know how many of you took the bet on how long it would take before I said the words AI. But we we spent a lot of time talking about the opportunity of AI in education.

And I think we've moved well past the novelty phase. Right? So much has happened in the last year that we've moved quickly into a heavy adoption, into a place where the majority of teachers, the majority of students are starting to think about the use of AI in the work that they do. Let me just do a quick poll. Raise your hand if any time in the last three months you've plugged something, anything, into a large language model, open AI, chat gpt, that sort of thing. Try to see this.

Alright. My super scientific analysis says eighty seven point three percent of you did this. I'm kidding. I'm kidding. But it was the majority of the room that had their hands raised.

And that and that's not surprising. Right? It's not surprising because we're seeing in our daily lives, AI is being embedded into it. There are generations of learners who are living with AI in the work that they do. But it's also interesting and it's also important because it creates an opportunity for us to address what are persistent challenges in education. Some of these challenges are newer challenges.

Right? Thinking about scaling as you get larger and larger class sizes, a thousand student class, a five thousand student class. Can AI help us solve that in a new way to allow us to scale? Some of these are persistent problems that have been talked about for decades. This idea of personalization. How do you get each individual learner? How do you make it an equitable experience for every single one of them? How do you use this to keep up with what is going to become a generational shift? Right? There is a generation of early learners that was born in an era of generative AI. There's generations of learners today who are actively using it.

And how do we solve for the needs of this modern learner? And then, of course, something that's probably real and near and dear to all of you is the reality of a resource constrained environment. Right? What do we do? How do we do more with less? But AI gives us an opportunity to think about solving these things. But what that's done, as exciting as all of that, is it's created this rush, a rush to provide new solutions and tools, solutions and tools that maybe start with the technology instead of starting with the problem. And it may feel confusing. It may feel confounding to you from time to time.

And I know it does because when I talk to you, I hear about some of these things. Right? Some of you have said, there's just so many tools out there. How do I even begin to figure this out? How do I solve really, really hard things? How can you help us solve really hard problems? I want to make the difficult things easy. How do we make sure this is an equitable thing? How do we make sure that whatever AI is doing creates an equitable playing field? How do we make sure that this is available to everyone? It doesn't cost too much. You don't have to make it so that only some can have access to it.

Some of you have institutions. You're far ahead. Right? You've got your own LLM models. You've said, hey, how can we plug this into Canvas in a more deep way? How do we make this more deeply embedded? How can you help us with that? And so as we thought about that, our approach to AI has to be different. Right? And I'm inspired by this quote.

The reason I say that is because I know that AI, like many of the technologies that came before it, has the potential to be seen as a replacement, a replacement for the human, a replacement for the educator. But that's not what it is, right? Education, education is about human connection. It is about the relationship between the learner and the educator. It's about that connection between the learner and another learner. It is about the connection between the learner and their advisor.

And at Instructure, we've always been committed to centering all of our innovation around the people, those people that we serve, those relationships. How do we ensure that we serve those relationships? And so our mission with AI, as with any innovative technology, is really about one thing, which is amplifying the impact of an educator. It's amplifying the impact an educator can have. And that is what has informed our AI strategies. So we've we've built it on a foundation of principles.

Right? We shared you principles last year. We built it on data governance, on knowledge building as the foundational layer. We've embraced what is an immediate opportunity for content generation. That's a really quick place to go to. Certainly things that we've done on our own, but we've leaned into a network of partners, of tools with proven efficacy and outcomes.

But if you go up a level from that, it's thinking about ways to accelerate use cases things that can meaningfully solve problems and workflows in ways that only we can uniquely do because we've got the context of the LMS, because we have the understanding of what that learning is happening. Leveling up on our history of being open to the next tier is making that capability available to you all to say, if you have a large language model, here's how you can plug it in and take advantage of our open ecosystem, our AIs around APIs our APIs around AI and making that available. And common through all of this is reinforcing equitability, making sure that whatever we do serves every learner equally, making sure that whatever we create enables it to be available to every learner out there. And so the result, we think, is going to create a superpower for educators. It's going to help them do what they do best, right? It's all about that connection helping them help their students succeed, creating time savings that allows them to use their time in higher quality and more meaningful ways, spending it where it matters most, and getting them faster access to education, helping you all narrow in on the things that matter most to you, and building these learning experiences that center right around that student.

So I'll stop blathering on about AI and talk about what that means in the innovation in our products. So I'd like to invite on stage Rachel Swanson to join me. She's going to help us show you how we're bringing some of these capabilities to our educators deeply embedded into Canvas, the LMS, through our products and through our partnerships. That's right. And while she's getting set up, I have a quick acknowledgment and a quick thank you.

The rapid rate of change of all of this technology means that the ground truth is that we are often learning and building right alongside you. Your feedback has been critical to us to make sure we're getting it right. This year, we launched a new program called Instructure Customer Discovery Sessions. Those of you who have been here and been customers for a long time, you may know this as Khakis from a long time ago. And we built that to invite customers to come in to work on proposed solutions, to help us improve what we were thinking about, to look at the things we were doing and give us feedback and challenges in some ways.

And we got a lot of challenges in places that help us sharpen our thinking. These multi day sessions would help us iterate, help us co create, help us prioritize jointly. And so some of what you're going to see today will be delivering on things that we previously talked about, but some of it has evolved thanks to your feedback. A lot of you have talked about focusing on the LMS and making the core LMS better. And so you'll see some of that.

And so all of this has only been possible because of a partnership from folks like you. So thank you for that. So with that, let's start taking a look, at some of these enhancements, shall we? And I think we're going to start with the place where you spend most of your time as teachers focusing on the teacher experience, which means we're gonna start in Canvas. Over to you, Rachel. Alright.

Thanks for having me, Sharon. You know, eight short years ago, I was sitting where all of you are at Camp Canvas. Who here was at Camp Canvas? Alright. So, you know, both as a customer and now an Instructure employee, I'm always so inspired by our latest innovation, so I feel really honored that I get to share these with you today. So we're going to start with the teacher journey by taking a look at, how we how we have the ability to create high quality content and evaluations, provide faster, more personalized feedback for our students, and also differentiate and provide additional resources for our students.

So we're going to start that journey today, by building a lesson for our students using our new block editor. So you'll see that I can, create a page from, scratch or with the use of predesigned page layouts using templates. But I really want to show you just how easy it is to create from scratch using the new block editor. So when I create from scratch, I can choose what elements I want to add to my page. Don't worry, we can always add more of these later.

And the color palettes that I have available are all fully accessible, as well as font pairings that are going to promote the proper use of screen reader technologies, and look at that, my page is built. I can very easily drag and drop these, elements around. You know, a lot of times we hear the need to create more engaging content, and so we can easily do that. As I add new content to this page, I'm going to add a single quiz question. I was expecting that.

And so now after I've added this quick check throughout this lesson, maybe I'm ready to add additional content below that quick check. And I'll go ahead and add another block of content. And now, of course, all of this is edible, I can rearrange these in the orders I want as the lesson comes together, But let's see what this is looking like from a student's perspective. I can go into preview mode, and of course I'm on a desktop device now, so I'm taking a look at this in desktop mode. We also know that our students are looking at content with a tablet as well as most frequently probably on their mobile devices and so we can take a look at that.

How fast and easy was that, Sharyn? Incredible. Block Editor is going to fundamentally change the way you create course content. Built into Canvas by default, it's obviously gonna speed up your workflows. It's gonna save you time in course design. But really importantly, it's going to build with accessibility in mind and it's going to account for instructional design best practices as it's built.

Okay. So some of you are probably aware that we just, released the first phase of our Rubrics Redesign initiative. So now that we've created a lesson, let's evaluate students with the latest rubric features. Here I have all of the rubrics in my course. I can sort my rubrics now.

I can also search for a rubric. Yeah. I also, can see where is that rubric currently being used. Alright. There we go.

And I also can edit, duplicate, archive this rubric if I need to. Alright. And now let's take a look to see just how easy it is, to work with building a rubric. It's really a new superior, user experience. So as I edit a criterion, if I want to add add that.

Alright, and now let's see what grading with rubrics will be like now. So, of course, we have the current traditional view that you'll be able to choose from, but we're now introducing a horizontal view as well as a vertical view. And you'll be able to make your selection in SpeedGrader on which rubric experience you prefer to grade with. Alright. So now that we've built some new content for our students, we've used new tools to evaluate students, Let's check-in on the engagement of our students, particularly with a discussion.

I'm going to use our new discussion summary to quickly summarize an ongoing discussion. So here you'll see I now have a new Summarize button. Go ahead and select that. And by selecting Summarize, I'm going to receive an AI generated summary of all the students' responses. If there's a particular area of the discussion that I really want to narrow in on, I can go ahead and do that.

This discussion was really related to some predictions students were making, so I really want to understand that specifically within their responses. So now I'm getting a summary that is specific just to the prediction students made. Discussion summaries are particularly useful in large courses with high enrollments. As you all know, often really great content can get buried in threads, and so as a teacher, discussion summaries are going to surface that great conversation to me without needing to dig for that. And also as the conversation continues, these summaries are just going to update as that that conversation, continues.

Wait. Wait a minute. Wait a minute, Rachel. This is only visible to teachers, right? Oh, yes. Great point.

Yes. Important detail. And Rachel, as an educator, I know you've never waited to the last possible second right before your class to go through and summarize the content for me discussions? I I would never do that. Well, in all seriousness, we know educators are time strapped and we think that this is going to be such an incredible time savings for you as it keeps you engaged in the conversation with your students. Alright.

So now that we have a pulse on the engagement of the class from a discussion perspective, we're gonna transition over to SpeedGrader, where we want to get a pulse on how individual students are doing and provide some feedback. We can now personalize video feedback even further with screen capture. Here Yes. Here I have a screen capture I recorded to point out specific strengths and opportunities to improve this submission. So previously in SpeedGrader, we could annotate student work and we could record video feedback, but with the release of screen capture, we can now bring both of those types of feedback together in one place.

So this is really going to open the doors for project based learning support as well as hands on activities that might require a little bit of additional explanation. And while we're here, we also added the ability to randomly sort students here in SpeedGrader. So remember those customer discovery sessions I was talking about earlier? This is all you. This is one of those things that came directly out of those customer feedback sessions. Yeah, that's right, Sharon.

And here's another quick one. You wanted additional criteria for messaging students, so we've expanded those options as well. So here I have a, message prepared to go to students who have not yet submitted to this, assignment, but now I can skip my excused, students. So this is really going to make communication far more relevant for students because I'm truly only receiving messages that apply to me. If I'm receiving this message and I had that, assignment excused, that wouldn't feel very relevant.

And so we have made those, updates for you, and a few others that you'll see throughout Message Students Who as well. Okay. So, as the most used LMS in higher ed and k twelve in North America, we're seeing more and more examples of dual enrollment or dual credit opportunities. So I have a couple of students in this class from my local high school that are enrolled, in my psychology course. I'd like to share some specific resources related to the dual enrollment program, just for those students, so I've actually created a module here for those students.

I now can go ahead and assign this, module specifically to some sections or specific students. And as you all know, the opportunities and the possibilities for differentiating modules for students is really endless. Yeah. That makes total sense. I I can imagine you seeing this being used for if you've got a set of students you want to personalize for students that need more support or the ones who you want to challenge a little bit more? Yeah.

Absolutely. You know, so, selective release of modules, I really feel like is one of the key updates we've made to personalizing, learning for students. And it's really common when we're personalizing learning for students that they need more than one resource. We need to put a collection of resources together. Often we think of those as playlists, and, this is really gonna enable us to do that now.

Amazing. Okay. So let's wrap up our time here in Canvas, by checking some messages we may have. Before I do that, another customer feedback request we had is you wanted out of office replies and signatures, and we have that now. Okay.

And as I return back to my inbox, it looks like I have an unread message here from a student. Using our new AI powered translation feature, Canvas is going to translate the message, identifying that my default language, is English and not Portuguese, which is the language that this, message was sent to me in. This is really an AI powered innovation that's going to bring a new level of inclusivity to our learning communities. So I'm gonna now respond back to this student. You'll notice that signature there.

So I have a response back to the student. And if I wanted to, I could also choose to include a translated version of this message to the student. So we are supporting over a hundred languages in our translation tool. And now when I send this message to the student, it will be, in Portuguese along with my English version below it. And lastly, translations are going to be available in both inbox, as we're seeing here, but also discussions.

So Rachel, I'm sure we've all seen some pretty strange automated translations. What makes this different? Yeah. We sure have. So this is really a best in class translation model. It's going beyond those typical, translation tools that we see.

It's taking into effect things like nuance, context, even figures of speech that wouldn't, translate literally. Wow. Amazing. So that does wrap up everything I wanted to share for our core Canvas improvements. We really just experienced some of the latest innovations and top requested features, the team has released over the past few months.

Incredible. There's really so much in here. And what you just saw was a sneak preview of some of the things, but there's a tremendous amount more. Those customer rediscovery sessions revealed sixty wins that we've put into the product. There's a ton additional, a ton of additional innovation in there.

But the key thing here is that everything that you saw just now, every single thing that has been worked on in Canvas, All of that, including the AI stuff, that is included with your Canvas license. Right? You get all of that for no additional cost. Everything you've seen today. So that's great for teachers, but we know that teachers are not the only ones who are time strapped. Right? We know administrators are time strapped too.

One of the biggest things they're looking for is a way to make sense of all of the things happening across their institutions. And they're looking for insights. And it's not that they don't have data. They've got lots of data. They've got data.

They've got data in data warehouses. They've got data in data lakes. They've got data in data lake houses. But finding that insight is difficult if you don't have a team of experts, if you don't have a team of analysts, if you don't have third party tools for visualization and reporting. And so we asked ourselves a question.

What if we could put that capability directly in the hands of educators? So over this past year, we've been working on some product innovation in partnership with a number of institutions, including San Diego State and Oxford Said, to co develop an offering to do just that. And the result is our newest product offering, Intelligent Insights, which is officially launching today. So Rachel, can you tell us what Intelligent Insights can do? Yeah. Absolutely. So Intelligent Insights is going to leverage AI and self-service analytics to solve a range of institutional needs.

We're gonna start with one of those needs, which is course readiness. So the insights in course readiness help us to ensure that all courses across our institution meet your standards for high quality courses. You know, as a former Canvas admin, I know the time that it took to ensure courses were ready for the semester across my organization. And often, the way we learn that courses aren't ready is help tickets from students. Right? Maybe they don't have access to their courses or key, critical elements of their courses.

So with course readiness, you'll be able to establish the criteria that indicates courses are ready according to your standard. In this case, we have three hundred and twelve courses that are not ready and a hundred and sixty seven courses that are. So to better understand what is contributing to this, let's take a look at this criteria chart here, and we can see that the things that my institution requires, the course needs to be published. That's a great first start, right? The announcements need to be published, the syllabus needs to be visible to students, and we're even asking that the word count of the syllabus syllabus be over twenty words, you know, ensuring that it's a syllabus of significance. I It sounds like this is something you relate to.

Assignments need to be published, new quizzes need to be published, and modules, need to be published. So I also have a disaggregated view of this, data in a table below, which is of course exportable. And not only do I have insight into the level of readiness, but with one action, I can send a message to all teachers whose courses are or are not ready to ensure that they are ready by the start of the semester. And, you know, sometimes we need to apply different criteria to different subsets of courses. So I'm going to go ahead and switch my criteria here, And you'll see now I have an entirely different set of criteria.

You know, maybe the nursing program has different set of criteria than the education department, and by using filters, I can then go ahead and apply to that subaccount in order to see exactly which courses are or are not ready, let's say, for that nursing program. So course readiness, insights is going to provide you fast access to your course readiness information, but also result in a ton of time. But ultimately, what this is leading to is a superior student experience because they're going to have what they need when they need it. But we also recognize that your learning management ecosystem spans far beyond, just Canvas, but also the third party tools that you have integrated inside of Canvas. And so we're gonna take a look at the Intelligent Insights, LTI usage dashboard.

I do have a caveat here that I want to say. We are looking at this next dashboard using a design tool, but that's only because, as you can imagine, LTI data is really hard to fake. We can fake a lot of data. LTI data is one of those hard ones. So, you know, as a former admin, I had a requirement that all of my third party tools be integrated in Canvas, which was a great student experience.

It was really hard for me, though, to manage that ecosystem. So the LTI dashboard is going to, show me all of the LTI tools across all across my instance, regardless of where they're installed. So, you know, it's really easy for us to figure out which ones are installed at the root account, but we know those LTI tools that sneak their way into sub accounts and courses, And so this is going to roll all of those tools up. I can see at a high level the details about the LTI usage at the top. And like the other dashboards we've seen, a table, underneath it with more granular data.

And then I'm also going to have two charts that are showing me trends in LTI usage over time. I'll also be able to filter, so I can filter by LTI, subaccount, course, role, DLTI version, so if we need to compare one point one or one point three, we have that option, and the date range. You know, filtering is really important as we try to narrow in on answers to specific questions. So I'm curious, what tools, teachers have launched in, the last thirty days. So I'll go ahead and make that selection and apply those filters.

So now I'm really narrowing my data in on that specific, use case. So my institution just purchased Canvas Studio, and I want to take a look at its usage and decide what to do next to further its adoption. I can look into specific metrics here in the table related to, what courses, are using this LTI tool, what subaccounts are using this tool, and even what unique users are using this tool. I know, it's pretty good stuff. And for those of you who use Impact, you would also be able to take immediate action on this data by launching a cam campaign to the users who have or have not launched this LTI tool.

And then when we look at the tool specifically, I'm going to have all of the same metrics except specific to this particular tool. So now that we've taken a look at, the engagement of our students, I'm sorry, now that we've taken a look at our content, let's take a look at the engagement of our students. So students in need of attention is going to allow institution admins to proactively monitor leading indicators of student success and engagement. Similar to course readiness, you'll be able to establish your institution's criteria for students you believe are at risk or require an intervention. Based on my institution's default criteria, I have fifty three students across twenty two courses that are in need of attention.

So we're going to consider both course engagement and course performance, so activity versus grades, in order to identify those students. The table below is going to, provide me a disaggregated view of the data above, which is, of course, downloadable as well. And like, what we've seen so far, all of the dashboards with Intelligent Insights are all filterable. So I can filter by things like subaccount, courses, teachers, the course status, and even a variety of activity criteria. So let's go ahead and change the criteria for students in need of attention, just like I did for our course readiness.

So, you know, it's often that, based on, use case, we might need to, change that criteria. So let's take our nursing program, for example. We might have a different criteria for our nursing students to remain eligible, And so I can change the, criteria there to be, maybe grades that are under eighty five percent, and I can apply this specifically just to my nursing program. And so now as I apply that, these are all of my nursing students that might require, to be reached out to based on their, current scores and status. And because we are really promoting, actionable data with intelligent insights, I can, with one single action, send a message out to all of those nursing students.

Okay. And so we're also launching, as a part of Intelligent Insights, a new Analytics Hub, But whether you're just using core Canvas Analytics, you're using Mastery Connect where you need quick access to your assessment data, or you choose to adopt intelligent insights, we wanted to provide you a centralized hub where you'll be able to access all your data and analytics across all of our products. So Rachel, that's all very exciting, very cool stuff, but I feel like there's got to be a way to, like, sprinkle some AI on this. You're absolutely right. So as our final stop, let's take a look at how we are using AI innovation, to allow account admins to ask and answer questions in natural language with Ask Your Data, the final component of Intelligent Insights.

So let's start with a question I would have wanted to know the answer to. And that is, which students have a course total below eighty five percent, but the teacher has not graded any submissions? Now why might you want to ask a question like that, Rachel? Well, you know, this is actually one of those early, early alert pieces of data. What I'm trying to understand is if students within the first couple of weeks of the semester are struggling due to the content or instructor support. And this is also going to help me identify, do those instructors need some additional support? And so the results that's going to yield for us is going to look something like this. Now if I wanted to, I can continue the conversation with Ask Your Data, But what is really, really cool here is I'm going to receive an AI summary, kinda summarizing the results that Ask your data found.

It's also going to share with me the methodology that it used in order to find that data within, your Canvas data. Yeah. Okay, but that's not really the end here, because now I have the ability to, select different data elements with the data, that resulted, and I can create visualizations based on that data. So I'm going to convert this table of data into a bar chart. We also will have options depending on the data that are a pie chart, a line chart, or even a scatter chart.

But now this might be important data that I want to return back to, so I'm going to pin that. You'll have the ability to create as many pin boards as you want to continue collecting, the insights that you are, finding with Ask Your Data. So I can go ahead and add that to my student performance, or course performance, Pinboard, wherever I'd like to add that. And as I come into my Pinboards, now I have a collection of visualizations and answers to questions that I have previously asked and saved. And if I want to then refer back to one, I can go back to that initial query, and I also have the ability to see the SQL that was used as a result of, gathering that data.

So, you know, while Ask Your Data is really intended to quickly find insights to your questions in natural language, It's also incredibly helpful for our developer users. It's going to help you speed up your rate of development because you'll be able to gain access to the query for things that you've asked in natural language. So we're really thrilled to be able to offer you your data in a more accessible format. Wonderful. Thank you very much, Rachel.