So there's, I think there's basically three three major components to to solving that problem. So there's first a content enablement engine. So we started with a no code platform that is called copilot designer. And what it does is allows non technical people to create immersive learning content very quickly. So the same type of content that we were creating back when we did that berry demo, would take five months, 10 to 12 people, and we cost a couple $100,000 To make that content. Now, a non developer can create that same content, same length, same type of engagement style, probably in seven days, with one or two people. And that's just like relentless optimization of workflow, and basically taking tasks that Unity does elegantly, but only a developer will know how to do and turning them into things that happen in background. So you just do a lot of background processing, of, you know, basically, and simplification of tasks on the front end, because getting your users very, very different. You know, we targeted at learning designers are just business people, people who are in the business unit and understand the problem that have the choice to either go create a PowerPoint deck, make a video, or potentially do you know, use something like this, to then solve their problem. So that was the first the first component of this tool was was content enablement. The second one was content distribution. Because the other problem we have in this space is our engines that we that we they don't get, we've always been fixed off, we've always been focused on real real time, real time engines. And that type of content is usually compiled into an executable, and executables are really hard to distribute in the enterprise space. That you know, they typically go through it process that procurement process and in a lot of clearing houses, to let that be distributed inside the enterprise world. And we went well, that's not going to work because we need to distribute things like, let's say 20 times a week, 50 times a week, like we're going to backlog it so fast. If we go that route, that it will just break. So we got to build, we got to build some player technology. And this was very similar to kind of the complications I was seeing in early digital VOD days, we had 100 file formats, you know, 16 different players, you know, play your applications that could some could play subtitles, some couldn't. Some could play multiple, multiple audio streams, some couldn't. Some could play this codec, some could buy that codec, it was a mess. And until that got solved, there was no digital video distribution ecosystem. It just didn't work. And so we said, Okay, well, let's let's focus on solving that. So we built a player for distribution that interprets our the content, and doesn't require it to be compiled. So now we've basically made a file format that can be can be played from the server and interpreted by the player. And then the last piece was, was probably the one that were a lot of that's plumbing, like it had to happen. And of course, there's a lot of cool innovation and a lot of a lot of really important things that we're I think we're solving there. But the thing that I think we're most passionate about, if you were to ask anybody a Talespin is is this idea of, of how we can get to a proof of skill, proof of skill, proof of skill. So the third leg on this the stool is is the analytics the dashboard and skill measurement. Because we're learning platform, right if I if you can't, if you can't measure it, then You know that you got a big hole? And so if you think about that historically and like the web one web two worlds, you know, if you're watching the video, you can measure did the learner watch all the way through? Did they answer my questions at the end of the the elearning questions at the end of the lesson, you know, did they attend the webinar? There's, there's some, there's some metrics, but they're generally not near as deep as what we can make possible with this new medium. Right. So here, I can put a learner into A into that Farmers Insurance example. I can watch how they move through the space, I can see what what do they do tasks, I can see what they investigate what they don't investigate. And I can build a scoring system that interprets that back to a proof of skill, meaning, pick, pick the skill, you know, in this case, maybe a basic investigation skill, did they achieve it or not. And it's it's software, it's it's, you know, at the end of day, it's not it's not a facilitator that's that's doing that, that's that's running a session, that then has to be the analysis of that learners. Job well done, not well done, is coming from that facilitator. And those facilitators are always different. And so the consistency of measurement in any other way you would try to accomplish this makes it nearly impossible to have a crew a true proof of skill, you have too much bias in the process. So the idea here is that if we can build a scalable content creation engine, a scalable distribution methodology, and a system for consistent scale measurement, then you get to this this new kind of paradigm for learning. And I think the the big change we're trying to affect is if you get to a proof of skill, the resume process changes the employment process changes, like, like a lot of things change in terms of how we work with each other and, and how quickly we can assemble teams and all that kind of stuff. And, and it comes a little bit back to like, how entertainment does it if you really think about it, like there's this, there's like almost as inherent proof of skill, based on crew positions based on the way that teams assemble a baseball complex the processes that they ought to engage in, like there's an inherent trust and proof of skill. That happens, you don't go through a six week interviewing process to join a crew on a production most time like it's like a spin up fast. You're in pre production for four weeks. That's it? Yeah. So it's really trying to get to that new foundation, that was the kind of aha moment back in 2018. And now here, we are coming up on five years later of just trying to get the plumbing in place.