 And I'm the head of growth for Meshkin Machine Learning. At Meshkin, we build intuitive and accessible tools for the new generation of machine learning engineers. The number of machine learning jobs piloted by companies is set to quadruple in the next two years. But all of this AI still needs human brains to build it. And there are nowhere near enough qualified. What this means is the majority of new machine learning engineers are actually software developers re-skilling on the job. And this is a problem, because machine learning is still a black box to most of these people. And it's very different to programming. But if we know that the majority of these new engineers are coming from this pool of former coders, then why aren't companies actually developing tools that mirror what they're used to? We asked over 50 AI companies, and they told us that they are using Slack in their day-to-day operations, just like most of us. We also found that the consequence of unoptimized training was about 10 hours a week per engineer. That's a billion-dollar problem globally on this one aspect of machine learning alone. So we set out to say that we're going to focus on eliminating this operational waste in order to help companies get their AI to market faster. Mishkin Interactive Machine Learning is a tool built for this new generation of engineers. It solves a series of complex machine learning problems, and it allows engineers to pilot their models remotely from their preferred Slack channel. We developed this in collaboration with our partners and clients who helped us beta test it before launching it to a wider market. Now, with Mishkin IML, engineers can actually see what Mishkin is automating for them, like blazing fast GPU optimization, or multiple model scheduling in any environment. But they can also have a plain English conversation with our bot, who gives them recommendations on how to boost the performance of their models, getting your AI to market faster. In a competitor matrix of ML frameworks and DevOps tools, our DPEC algorithms are accelerating the machine learning on the back end. While on the front end, we are the only ones doing what we are calling chat ops for AI. Our IML works on a freemium model similar to what customers are already experiencing on Slack. With this and other tools that we're developing, we hope to capture 20% of this new ML market by 2021. And we are actually seeking investors and partners right now to help us with this global takeover. Our team has over 16 years experience in machine learning and deep tech product development and sales. Our investors include Listo Silesma, chairman of Nokia, among other top Nordic VCs. They believe in our vision to be a global leader in tools for this new generation of machine learning. Thank you. That's fine. Thanks. So Ray, I didn't get exactly what it is. So you use external machine learning platforms, like Studio ML. You do your analysis. And then you use this tool to be able to share and to communicate with other people within your team some of the results or, I mean, it gets... Yes, that's correct. You can deploy in whatever environment you're using, so Keras, PyTorch, it's platform agnostic in that regard. What this actually does is it has a daemon that's watching your models perform over time and it's giving you these alerts based on what it sees. And one of the things that it actually can help with when I was mentioning recommendations to boost performance, so things like hyperparameter tweaks, tweaking the batch sizes of the data, simple things that we can recognize, basic hygiene to clean up your models, which is actually boosting performance up to 100 times, depending on how badly things are going. Okay, and then you can communicate with your colleagues as well as with a bot, which also gives you recommendations based on that. Yes, absolutely. So one of the, I think, I don't know if I can go back to my slides here, but on one of the screens, you could actually see that our colleague, Edan's model, was performing poorly and the bot was recommending certain actions and actually another one of our colleagues was commenting on that and saying no and actually being able to control it that way. And what's your go-to-market strategy? So how do you find users for your product? So we actually are in partnership now with Slack and with AWS. AWS, we were just at their conference at Reinvent in Vegas last week, so that's actually where we launched this product. And our partnership with AWS we're helping their SageMaker users use our API to get on board with Slack. So we're trying to go through the partnership channel for growth right now. Yeah, thank you very much.