 Thanks time for the introduction. Yes, so as you've heard, I am at Curtin University working at the Curtin Institute for Computation. And I am overseeing the ADEX work we're doing from Curtin. I hope you all can see my slides and that things will move forward when I click my buttons. Okay, so quick introduction to who ADEX are. So our vision is to provide astronomy focused training, support and expertise to allow astronomers to maximize the scientific return from data and computing infrastructure. And we commenced operations in March 2017. This was after institutions from around the country could apply to actually run this initiative. We now have two notes, Swinburne University and Curtin University. And we're funded by Astronomy Australia Limited through the National Collective Research Infrastructure Ingress. If you want to have a look at our web page, feel free. And with the work that we're doing, we have basically three service components that we are looking after. There's the training. We do face-to-face training, webinars, internships, a lot of different kind of training. The face-to-face can be single or multi-day workshops. The webinars or pre-recorded tutorials are also available on our web page. And we're also running internships out of the Swinburne note that are there to support a lot of PhD students in this case, to actually learn about software development and work in a software dev team. Our second service component is computing and data services. So there we support, we do software support on the OSTAR supercomputer also at the Swinburne note. And data management and collaboration platform is also run from that note. And then our third service component is national support, where we do spend a lot of time on as well. And there we have professional software support, so researchers can apply for time. And we match them up with a software developer to work through their project and give them back a prototype at the end. And there's also the Astronomy Super Computing Time Allocation Committee. So as part of ADEX, we have a specific number of CPU hours that can be applied for and we got a time allocation committee for that. So to look more in detail at the training that we're doing. So the training we provide, we aim to teach basic computational skills and best practices. Personally, I do believe that as researchers, you do not have to be software developers, but you still need the basics right to make sure that your analysis works properly, but also to then be able to talk to the software developers, for example, that we have in our team to enable better work there. We are aiming to cater for different skill levels and we're currently at the time where we are going more into the intermediate and advanced training. We want to offer content and new computational advances that could prove really useful in the future, either for astronomy itself or just for data analysis in general, because the fact is a lot of PhD students don't stay in astronomy, they go out into industry or other fields. So for example, we've run quite a few machine learning workshops over the years as well with a focus on astronomy to talk with people about how they potentially could apply something like this. And we also would like to prepare researchers for alternative career paths within astronomy or the technical industry. So that's why we have the internships for example. Now, when we started out, the main question was like, what should we teach? So the first thing we did in 2017 was put together a community survey which had a range of multiple choice questions to just gauge what software are researchers using at the moment and what are the training requirements coming out of that. So in the last 10 years or so, Python has become really popular with astronomy. Before that you would have had a lot of people using IDL or Fortran or something like that. So the question there was having really bridged the gap yet, how much basic training do we still need in that area. So you can see from the graphs at the bottom that the majority of people are using Python, it's about 86% or so. We also have a fraction of people that are using R, I was one of them, so I didn't really start using Python much until after I finished my PhD. But you can see there as well that there's a lot of people still using IDL. And one of the problems in the way of IDL as well is that you need your licenses and institutes often have very limited number of licenses and you end up in fights about who can do the analysis and who can't. So getting them onto open source software was something that was important to us. So you can see as well there from the requested training that most people really wanted to learn more about Python and machine learning, which was up and coming in 2017 as well. And then how to do scientific visualization. I mean, we all need to present our data somehow, but often we do end up probably choosing the worst possible graphs to show the data. With the new influx of big data that we have coming in with new telescopes like the square kilometer array. We really need to learn how to migrate things onto high performance computing, how to make our code more efficient and also go into GPU coding where it makes sense. So these are also quite popular requests we had in that original community survey. So based on that we then actually put together quite a lot of workshops. We repurposed material where possible. So we heavily leaned on Carpentries material, both for content, but also for the teaching. So I myself am a Carpentries instructor. So that was quite helpful to know how to put the material together and to teach it. So the collection you can see here as well. The 23 things I don't know how many of you can remember it I repurpose that for 10 astronomy things to make that available to just talk about data management because as I said with all the big data we have coming in we really have to think about this much more in detail. We have the ASPO's that's the Australian Virtual Sky Observatory and they do a lot of that data management for us now. But there's still a lot of training out there that people would need on how to use those observatories properly, how to get the data they're actually interested in. The main languages we told was Python and version control and I honestly was surprised by how many astronomers didn't have version control background when I started my training. So I think at least in that sense, we have helped the community to actually put things under version control. And then we did a lot of training as well around HPC. So we used to have POSI as one of our partners as well, who helped a lot with the HPC training and getting things on to like their cloud for example as well. Otherwise we are, we have done a little bit of R and SQL training, and we have information on our website on HDF5 and specific astronomy packages in Python. This one is called AstroPy, this little snake you can see there. But the issue in a way is a lot of this is still very focused on introductory material and we've been thinking about for a while how can we actually get more into the intermediate and advanced space. So training delivery also we have tried to span as many possible pathways as possible. So we got our face to face workshops. Obviously this year there haven't been that many. We managed to squeeze one in just before everything locked down. We also got an LMS that is now quite easily accessible on our webpage where we have video tutorials and we're working through a lot of that to also give Jupyter notebooks along with that so people can actually work through things. A lot of the video tutorials in our LMS are also on our YouTube channel. All the materials from our workshops are on our GitHub page. As I mentioned earlier, we run internships as well at the Swinburne node. We also try to put together online resources from other places that might be useful for astronomers. So a lot of the online resources at the moment is a work in progress, but in the end I want to make this kind of like a little one-stop shop. So our face to face training is really where I put a lot of my time in. As I mentioned earlier, we adopt the carpentry style called a long setup, which has been quite popular and successful. We run single multi-day workshops. And I think most importantly, we really work together with the astronomy community by working with the Astronomical Society of Australia through their theoretical chapter, Anita, and the Hollywood School, which is run by students for students. Back when I went to the Hollywood School, it was mostly talks and for years people have said we would like more hands-on parts. So it's really good that for the last three years or so we've gone along as ADEX and have taught some introductory computing. And we've also worked with some of the centres of excellence like Astro3D or OSGRAPH. And this year we had our first hack week. So that was the last face to face training we had. And the hack weeks are really interesting. They're kind of a mix between a summer school and a busy week, but they're very participant led. And it was important for me that people come along, bring their own hack project that they can then actually use the skills they learn into tutorials and apply them to the project by still having people around that can help along when they eat it. Yeah, this is just to show that the material we have, we have been working on making that easier to access as well. So if you go to our webpage, you can actually see a summary of all the workshops that we've run and all the different kind of topics that they had. The LMS you can join the courses. It used to be behind another lock-in. Now it's right in front. So it's quite easy for people to go and look at things. YouTube channel for those who prefer watching things on YouTube. As I said, everything is on GitHub. We have a few success stories from recent internships on there as well and our online resources page is a work in progress. But the idea really there is to make it a one-stop shop for PhD students, but also researchers to just come and start looking into, okay, how do I actually learn what I want. So the question for us right now is where to from here. So how can we ensure the training stays relevant and how can we provide this intermediate and advanced training. So one of the things we started doing is collect skill profiles to figure out what people at different levels actually know in terms of computing. And then from there, map the learner journeys so we can identify the obvious gaps in the training material that's out there and hopefully fill those gaps and then offer easy pathways to follow to upskill to those steps. So our webpage right now for some of the core skills we are giving a little bit of information on what beginner intermediate advanced might look like. And we also got a couple of astronomer profiles together in terms of as a cosmologist cosmologist. This is the typical skills you might need. So if you come in as a new PhD student or so you can be like, okay, I need those things so I can start learning Python and I can start learning these packages. And the next thing we're doing right now. So we have a merit allocation program that until now was for software support. But we found even with software support applications that we received previously that people would want some training or they requested software support where it's like, well, it's better if you actually train on how to do it yourself then write those pieces of code for you. So we have worked on adapting our merit allocation process to also include people being able to apply for training. So this helps us understand what the training needs of the community are and you might find more interesting areas that are interested this way than if you send out another community survey. And it allows the community to request that bespoke training that they need for advancing their own research. So what can people apply for? They can apply for workshops either to rerun stuff we've done before or to request a new one. Current way they can either be face to face or live streamed. They can request online resources, so new self-paced tutorials or webinars to be created. And they can also request training as part of an actual software support project. So as I said, at the end of a project, we usually hand the code over to the researchers and from there they on their own. So here we made it possible to say, okay, don't just give us the code, put some extra time in to show us how to use it and how we can maintain it in the future ourselves. So we currently finished the EOI process for the next semester, which is the 2021 semester. And we had three training applications and we also had several software support projects that would like more training at the end. So hopefully we see what our time allocation committee is deciding, but hopefully we get all of these through and it's going to be very interesting to see how this continues in the future. So with that, thank you and we can have a look.