Data Science Hangout featuring all of us!
videoimage: thumbnail.jpg
Transcript#
This transcript was generated automatically and may contain errors.
Then why don't we get started here, so happy Thursday everybody, welcome to the Data Science Hangout and hope everybody's having a great week. If we have never met before, I'm Rachel Dempsey. I lead our Pro Community and lead up Customer Marketing at Posit. The Data Science Hangout is our open space to chat about data science leadership, questions you're facing, and getting to hear what's going on in the world of data across different industries.
But at the Data Science Hangout, we're all dedicated to making this a welcoming environment for everybody. We love to hear from everyone, no matter your years of experience, titles, industry, or even languages that you work in. So it's totally okay to just listen in if you want. Maybe you're doing some work on the side and we're just in the background, that's okay.
But there's also three ways you can jump in and ask questions or provide your own perspective. So for anybody who hasn't been here before, those three ways are, you can jump in by raising your hand on Zoom, you can put questions in the Zoom chat, and always feel free to put a little star next to that question if it's something you want me to read instead, maybe you're in a coffee shop or something.
And then we also have a Slido link where you can ask questions anonymously. And so we're doing things a little bit different today, and we're ready for it. We've done this a few times here. But we actually don't have one featured leader today. We are all the featured leaders today.
So this will be an open conversation with all of us. And we can all ask each other questions to the broader audience, kind of jump in and share our experience too. But I'll get us going and kind of kick us off on a topic. But if there's something that's top of mind for you, definitely share it into the chat here too.
Presentation tips for conferences
But something I've been thinking a lot about lately is of course, presentations because the POSIT conference is next week. And I know there's been a lot of preparation happening in terms of talks and slides.
So I thought that could be a fun starting point for us to hear what lessons have you all maybe learned along the way in giving conference talks or even presentations at your own company. And I'm just curious if anybody has some tips or things to share with us.
So I was just kind of starting off the conversation with everybody about presentations because POSIT conference is top of mind for me with it happening next week. And to just understand if anybody had any tips or lessons learned in giving conference talks or the way that you do your slide preparation as well.
Okay, Abigail, I might call on you first because I just read yours first. Is that okay?
So look, just this whole approach of like, you know, look, I'm a nerd. I like coding. I want to show people code, but like that's just never the answer. I mean, like you think it's going to be the answer. So really, it's like reverse engineering, what is going to be useful to my audience, you know, and not treating this as like this is an opportunity for me to talk, but treating this as like, what is it that's actually going to be useful in somebody's life.
Yeah, thanks. So yeah, Abigail's advice is absolutely spot on. But we've had articulation training. So this is articulation as a consultancy group. We've been working with Posit to help the speakers kind of get ready for PositConf. And the advice has been absolutely fantastic. I've fundamentally changed what my slide deck and presentation looks like from before I started.
What they got us to think about is like, what is the hook? So what is the thing that you're going to try and draw people's attention in? You know, because we're all dreadfully distracted by everything in the room.
And so you want to hook people in. And then what are, if you like, three key items that you're going to want them to remember later? And then once you've got those things, and you know, you may or may not have a call to action at the end of the presentation.
But once you know, what is your hook? So what is the key theme? And then what are the three kind of things that you want people to remember afterwards? Only then kind of flesh out each of those three. And then think about what is the visuals or the, you know, cues, code, images, whatever you want that support those three elements.
Whereas if you step away from PowerPoint or Keynote or whatever it is you're using, or even Quarto, and just think about what are those things that I've just articulated then. And I actually, I used a mind map tool, which allows me to think about, you know, what are those things and then explore each of those and go, what are the topics or what things must I remember to say about those?
Yeah, so full disclosure, I used to be an educator. So I have a lot of experience with keeping people engaged, especially when there's a lot going on. You know, the conference is busy. Some people are there virtually and you want them to listen to what you're saying and focused on you.
So definitely as few words as possible on the slide. And something that our now hybrid world has given us the opportunity to do is put a code or a link on your first slide to your slides and your notes. So you can tell people, hey, this is available to you.
So you don't need to be frantically writing notes if you don't need to. Everything's in there. Why don't we just talk and then keep it very, very simple. And that's going to help people focus on what you're saying more when they have that relief. Like, I don't need to write everything down. This is available to me.
Yes. So this is a earth shattering, but it's easier to remember. And it's more interesting to listen to when your presentation has a story. So just to remind people when they think about preparing their presentation, they can share a story because that will capture attention and people enjoy it more.
Yeah, I have a really hard time when someone puts just a wall of text up. It's a little hard to follow. So yeah, I think bolding is really helpful too. So not just change a color, but make it bold.
I found that the big images for a slide really helped me remember a presentation too. Like, even my own talk last year about community building, I was thinking, like, what were my four main points? And in my head, I pictured those four, like, the big images on a slide. And then I'm like, okay, I can remember.
Internal presentations and sending slides ahead
One topic, or one follow-up question on the topic of presentations, I was curious for you all, when you're putting together slides for internal presentations, do you ever send that information out ahead of time, like, or with a summary before a meeting? Or is it usually in the meeting that's the first time? People see it. I never know sometimes, like, sending something out ahead of a meeting, are people actually even going to go through it, or do you assume that they will?
Storytelling streamline talk that like you would do at this conference is very different from the kind of slide decks, I tend to send out at work, or that gets sent around at work, the ones that work. They're information dense, they have a lot of numbers, a lot of tables, a lot of charts, like, you're intending to spend time with it and read it and stare at the numbers and understand the report out of the results of the test or whatever.
So, I think it's super important for, like, information dense decks that are kind of half a document to be read and half to be talked through and as opposed to, like, this sort of conference where I think the decks will be considerably less information does, I guess I'm trying to also figure out what's the difference then and presenting a story to people in a conference and getting them to understand what you're saying versus, like, getting your colleagues to understand the key story behind the data too.
Yeah, I've been thinking about that because, like, I kind of wonder how useful, like, we have an internal conference here for analysts and I kind of was telling some folks, you know, maybe something like articulation ink coming to us before the next internal conference would be good. But that's because for that conference, a lot of people reporting out of their area of expertise, like, it's like people like teams are talking to teams that don't usually work on them.
Like, if it's your own team, and you're just kind of, like, people already know what you're kind of heard what you've been talking about for a couple of months, and then you're just showing, like, well, here's what happened with the test. I think it can be more information dense. So I think it really depends on the audience and how familiar they are with stuff before him.
Yeah, so, and I apologize for the background noise. I like to keep my slides very light. I like to use lots of images, use the 5C color and whatnot. But when you're meeting leadership, you have a very short amount of time to convey a bunch of information. So, what I usually do is take all of the images from my slide deck and put them into a sales slip. It's usually 4 or 5 pages, but it has all of the detail that you're going to cover in the presentation.
So, when you send out both, now they can see the slide deck. They can also see all of the detail that you're going to be providing, all of the backup and all of the narrative prior to the meeting. So then they come in prepared and they can ask really good questions. For me, it's been pretty successful.
Dealing with presentation nerves
So, I think, like, I've been through so many presentations myself giving them, but every time I give a presentation, there's always those jitters that I have. And I kind of wanted to hear, like, out of curiosity, how would, like, how do you personally deal with those, like, presentation jitters that you get right before it or even during it?
I took some training, actually, when I worked at Health Net giving a presentation to our execs. And what was helpful was really the power poses that they had us do, which are, like, so, like, weird when you start doing them.
Yeah, or something like that, like, power, you know, get in your chair, get all, like, make your body big. You do, there's, like, different ones that I'm sure many of you have heard of those, right, and you do them. But for me, that helps a lot, kind of get rid of jitters.
There's something like, it's almost like doing yoga in a way, because, like, you're stretching your body, you're breathing, getting the blood flowing. But aside from that, the other thing that kind of calms my jitters is rehearsing. Like, rehearsing it to death.
So even if it's just the five-minute speech or 15, 20-minute speech, I rehearse it and refine it to where I don't need a lot of words. But if I need, if I have the extra time, you know, I certainly can't fill it.
So, another thing is, I think it's natural to feel nervous. I think everybody feels nervous. So, you know, if you don't, then that's possibly a different sign.
But the other thing is that if you have practiced, right, then as you're getting up to do the actual performance, as it were, remember back that, you know, before you did all that practice, things were a lot worse. Right. So, you've come a long way since that first time you even thought of doing this presentation to now. So, if you've done the practice and you've got that far, you're good to go.
You've come a long way since that first time you even thought of doing this presentation to now. So, if you've done the practice and you've got that far, you're good to go.
This is a trick that's helped me because I read it somewhere that if you're chewing gum right before your big presentation, it kind of tricks your brain that because you're chewing something, you're having lunch, like, nothing must be wrong. You're just like eating like normal. And maybe it's placebo effect. Maybe it's real, but it has to help me every time.
Building slides in Quarto
But I saw Luke, you had a question about Quarto in the chat. You want to jump in? I think it is about if people are building slides in Quarto.
But the question Luke's question was curious how many presenters are building slides in Quarto?
I know personally I've used it a few times, and I've enjoyed it. But when it came to something I had to have a animation in, I had to go back to PowerPoint because I'm not quite that advanced yet.
But I've also been lucky. I go to the Southern California art user group, and Emil happens to be there and does his slide. He's done his SlideCraft posts as well as he's done the presentation on SlideCraft a few times, and they're really good.
Successful onboarding stories
But another question that I'm going in a completely different direction here, but I know something that's come up a lot in past Hangouts is around onboarding. And it's usually talked about in more of a negative way of companies doing a bad job of onboarding.
And I was just curious to learn from you all, like, what are some of your favorite successful onboarding stories and what made it so great? Because I think there's probably something that we all could learn from hearing those stories.
Like, I've never felt like I was always having to go at it alone. And then you just kind of gradually transition from, you know, new to semi-experienced or seasoned that then you can pass that on to the next person kind of coming in.
So I think one of my favorite personal onboarding experiences, my boss handed me this sheet of paper and he was like, okay, you know, tell me your preferred style of being approached. Tell me what your pet peeves are. You know, all of, like, tell me what you would have wanted to tell your last boss that ticked you off.
So having those kind of candid conversations with people is really important. And then when I got to Peloton, one of the things that our enterprise data team does is creates a user manual. So within somebody's, you know, first couple weeks, they go, okay, here's what makes me tick. Here's, you know, all these things about me and how to approach me, what motivates me, what am I afraid of, what am I really trying to do here.
And we love it. We still use that practice today and it really helps people kind of break out of their shell and make connections when we've got such a large team that's, like, you know, 40 plus people. It really, you know, kind of breaks the ice there.
Yeah, I mean, the job I'm in currently, when I first joined, it was pretty clear, like, hey, like, already when you join, here's the plan for 30, 60, 90 days. You know, we have, like, a project for you coming in and we expect, like, this amount of progress and we'll check in.
You know, and that will help you both get your feet wet, but then also provide, like, sort of this clear direction rather than, you know, just ambiguous, like, sit in on meetings and, like, pitch in where you can is very, you'll get up to speed at some point with that method. But it's just a lot quicker, I feel like, with that, like, plan ahead of time from the hiring manager.
Oh, yeah. I was just going to say, a lot of times, in my experience when you're onboarding, people seem to still be, like, they're, like, worried you're not up to date on the skills, but they've just vetted you on your skills. So, the best thing you can do for somebody is the specifics, like, this is how we connect with our server and this is how you, you know, the specifics to your area, you know, we have this platform here that requires this funky code.
And then the other thing is that we're doing in my current position is we have markdowns, notebooks, whatever, for all of our meetings that have from all the way from like here's how you access the spreadsheet we put it in every single one. So, a lot of onboarding has the opportunity to just go through this stack of notebooks or whatever, and see how people have done things in the past. And that a lot of times those little details are actually in there.
Another thing that I get better about when I'm helping onboard somebody or just doing an intro call with them is just reminding myself that they're being thrown so much information right now and they're not going to retain every single piece of information. So, if I could be the person to actually like tell them a story about a customer or something that they're able to remember, rather than like just throwing more facts at them and giving them more time to like come ask about the facts, but to at least like understand the culture by the story that I'm telling.
Yeah, I completely agree with the idea like when we onboard new data scientists, you know, there's, I kind of assume that they know Python, they know Pandas, they know like how to do, how to be a data scientist, but they don't know all the quirks of where our data is hidden, like how the weird Parquet files that we use, not weird, but the unique Parquet files that we use to store our data, and how it's organized by date tables.
So, we have some very formal SQL training that introduces people to like our kind of enterprise, you know, SQL type tables, but then I have written some Jupyter notebooks that are very much, I mean, they don't, they're very lead by the nose. They just basically show, and so something they can refer back to later, like, how do I load this kind of data? Oh, here it is. I did it in this notebook, right? And so it kind of leads them through step by step, like loading some data, making a toy model, like using some of the quirky tools that we built internally. And I do think that is the most important thing, at least for new data people is, yeah, we know they can like understand trends and they can do explore data analysis, but like, what are the quirks of our system and how can we best help a new hire understand those quirks?
Hi. Yeah. So a couple of comments, right, that depends on the size of the organization. You know, typically if you're like an enterprise, there's, there's an onboarding team, right, you really don't have to worry about that. I mean, earlier, there was a comment made on markdown files, right. Yes, definitely documentation right to be able to onboard those team members. But what I found more important is being onboarded even to the culture of the, of the organization.
And so now I get it. Had I had a mentor that could have said, Oh yeah, there's more than enough. So, I think having or pairing a new hire you with a seasoned employee, a mentor helps tremendously, and to onboarding to the culture because all the technical documentation, we're all smart here, you could read, you know you can move forward to it. If there's a hiccup, the mentor will be able to support you there so I think that's my big take on on onboarding, definitely a mentor.
Yeah, we have a four week onboarding at my where I work, and it's like drinking from a firehose, it's where does data come from data privacy, it's the tools we use because it's a variety of tools. So I think something that's been really helpful is I added a analytics workflow wrap up. And so it's just like, hey, these nuggets that we mentioned that you might have missed are like very important.
And so it just kind of is a thing that I think they can reference for a while after onboarding that's kind of how I present it to them is, you know, here's a slide deck, but also, like, keep it, keep it handy because there's a lot of little things that would make your life a lot easier. So after they've gone through the data come through and all that other stuff, we just do this wrap up to connect the dots.
So I think, like, having a good understanding of, of, like, the positions that you're probably helping or in departments that you're probably helping with, with their data is just getting to know what are they doing, what are the things that are important, important to them. And then also, it helps with understanding why the terminology they're using doesn't make any sense to you, or why specific data terminology to you just doesn't make any sense to them and just just helping bridge that sort of gap and really getting a better, better overall experience about what this company is doing, what your goals are and how you can think about some stuff that isn't kind of directly in front of you as well.
Interviewing and hiring practices
So we've had a couple of job interviews out. So I work in the NHS in England, and we've had quite a number of them. So it's taken a lot of brain power to get through all the application forms. But what I would like to do is just check with people what kind of experiences that you've had that are really good to use. So I've mentioned about tests. We've all kind of agreed from ourselves that tests didn't really work out. And does it really show somebody's very good.
So it's just having that idea. We are restricted by some of the things we have to do. Like we must always ask the same question of every single candidate because it's part of the public sector approach that we have. And we'll have the same amount of time pretty much unless people have some extra requirements. But it was just ideas from people. It'll all be online as well. So, you know, maybe things like questions, because sometimes they can be quite long, can go into the chat, for example.
So, I said that I really enjoy it when I can tell that the interviews that I'm in are focused on me as a person and who I am and if I'm going to fit and if I'm a good human being versus my skills because you can get anybody in there with the skills. Making me feel like who I am matters to the role and that my soft skills are also valued has been really important for me because I'm a technical person who has great soft skills. So, hearing that those are valued is pretty great.
And recently, I had a really good interview where I had sent questions ahead of time. I think I had sent four questions. I do that a lot. I send things ahead of time as much as possible to help people be prepared for things. And then the interviewer just took my questions and let them frame the interview. Like, I will just go through and answer your questions and you can answer back to me with counterpoints or whatever from your experience. And that worked really, really well and made me feel really welcome.
Yeah, I've been through some processes recently for the US federal government. So, similar constraints about how fluid you can be. But there is a pilot that they're doing called the SME QA process. And so, it's subject matter expert driven. And as part of that process, they allow you to do kind of a use case and provide a coding sample, which is a great way to kind of highlight your skills. Because normally, it's like you said, being in the public sector, you're very constrained and don't get to kind of show your individuality and skill set.
Building and growing user groups
Well, I've been listening to you all for a little bit and just while I do other work. Yeah, so I am at Tulane University. I'm a graduate student in the Ecology and Veterinary Biology Department. And the group has been around for a little bit, but it had a COVID hiatus, and now we've rebranded, and we call ourselves TWERK.
And it's very New Orleans, Tulaneans who enjoy our coding. And we're just, I'm just one of the kind of collaborators, organizers, and trying to, you know, workshop ideas on, like, how to get people involved, kind of, I guess, what are the best, like, onboarding strategies. It's like, hey, bringing people in and, like, getting them interested or comfortable. Some folks come and they're like, a lot of graduate students have a crash course in R or, like, a rough relationship with R when they don't have, like, a more coding background.
Yeah, absolutely. So I can speak to, I lead, I work for Charles Schwab, and I lead the data science at Schwab user group. Our last meeting, we're actually trying to schedule something with Rachel and pause it pretty soon for our next meeting, but our last meeting, I think we had a little over 200, 230 or so users from across Schwab come in and hang out.
But I'll just echo, like, it hasn't always been that way. We started out really small. We started out as an R user group, just purely R, moved to R and Python, then moved to data science. And when we were the R user group, it was basically the same five people who were just getting together monthly and presenting back and forth to each other.
And I came in with zero R knowledge, too, right? I was a SAS user, SQL user at the time, and wanting to learn R. And so everything was really cool to me, right? I mean, they're showing me, like, basics of dplyr, and I thought it was awesome.
But we started small. I'd say a couple of things that we've always tried to emphasize, right, is that all levels of skills and ability are welcome. So we've got some folks who come in and bring neural networks and self-learning maps and just, like, really complicated financial topics. You know, we've got PhDs presenting back to us, you know, the specifics of which are probably over a lot of people's heads, but everyone finds it really interesting.
But we also encourage the person who, like, just learned how to do something cool in, you know, in R and dplyr and Python, whatever it is, right, if you've just learned the basics. We had somebody who was, you know, I was just posting about something I was pretty proud of. We had a user who figured out how to connect to our big data environment via Python rather than using the console. Like, those are just small wins. So, like, to include everyone, I think, is a really good chance, plus giving everyone a chance to present, too.
So, that's really helped, and I remember five years ago when we started this, the R and Python community here was just so fractured and so small, and you really had to, like, I mean, it was basically Google and Stack Overflow to figure out your answers. It still is to some extent, but now we have, you know, 200-something people who come together at least once a month, plus team channels and stuff, and we're all helping each other and deploying packages that we can all share and things like that. So, I just encourage you to keep it going.
On the community topic, too, I feel like something helpful to just add or say, too, is that I know, like, we do have a tendency to, like, celebrate as things get to be, like, really large groups, and that's amazing, too, when we have a ton of people engaged. But also, even if you have a small group of people coming together once a month, that's awesome, too.
So it doesn't always have to be, like, a huge group, too. Like, we have a data science community builder coffee hour that we do once a month, and there's about maybe seven of us that will join pretty frequently, and we always have great conversations there, too.
And that's been something I have to remind myself. Like, it doesn't have to be a huge group to be a successful community, too.
I think part of it, and I'll talk about this at PositConf, is really making it easy for new members to contribute. And I think there's a lot of different models you can do that. And I'll talk about what we do in the Portland R User Group, which is everyone kind of brings in their problems, and it kind of normalizes that you're having problems, that problems are okay, and that you don't need to know everything.
Finding mentors and peer support
Oh yeah just sometimes I get really in the weeds of some niche issue and I just really wish someone could just explain to me what I'm missing. And it seems to be out of scope for some of the places it's not out of scope but you know I'll put a stack of a button will just sit for a month or, you know, I kind of want to fix it by end of day and I can see in my mind like where I don't know I just wish I had someone I could tap on their shoulder and say, What's going on here.
I have been doing something recently that has been super super helpful Jacob, Jake, which is just having working sessions in fact I have one later today with a friend. I have been weekly just getting on one on one with somebody for an hour or 90 minutes and working on stuff.
And there's, it's so helpful to just be able to talk something through with somebody and be like I'm trying to do this and this isn't working and sometimes somebody will say oh have you heard of this package and I've never heard of it before. And just being able to have a mentor peer mentoring has been really valuable for me so I would encourage people. Yeah, it's not always somebody who has 10 years of experience over me sometimes it's somebody who has less experience than I do who's like, Oh, have you thought of this super basic solution and maybe I was being too complicated in my head.
Just really quickly I wanted to agree completely with what Libby just said because, especially during the early pandemic co working was like, awesome, just because I learned so much from people who were outside of my field they were just friends across the country and we would just be co working just working and then you can shoot a question out or I wonder about this or whatever it also helps keep you motivated.
But I wanted to, it's funny I wanted to echo something that we have said recently outside of this chat that instead of necessarily thinking about oh I need a mentor for this or I need, you know, this type of person. Instead of focusing more on just building friends, like making friends in the community that you can then just text.
Instead of focusing more on just building friends, like making friends in the community that you can then just text. And it's really really difficult to do sometimes especially because I know I'm an introvert, but just having making more friends that you can text in that way is such a silly way to think of it but it, it can really help.
So it's kind of echoing really what people have said in the chat and also Amanda in answer to Jake takes a question. I'm part of the NHSR community and we have a Slack group. And so it's part of that you know that you have a community when other people answer your questions for you and even when I ask I get people answering and so my mentoring that I've had has been from whoever answers sometimes.
And although I don't know people specifically and I think it does hit that introversion level, we chat a lot through text, but I've never actually spoken to these necessarily people face to face. But I know that sometimes if I put a question out, certain people will latch on to these ones because they're so good at that particular area.
And also what's really great is when somebody else says, I can't solve it either. This is a big problem because you said like when you put it on Stack Overflow and then nobody answers for months and you're like, is that just because I haven't worded it or is everybody really this lost? It's really heartening to have a couple of people go, yeah, this just doesn't work.
What's happening? And then at least you've kind of commiserated together and you're not alone anymore. So the mentoring has never really been one to one. But as you say, if you find friends, then you can go to people and say, can you help me with this? Particularly, particularly if it kind of strays into more career focus as opposed to problem focus.
Data governance and APIs
Oh, yeah, thanks. Yeah, data governance. Yeah, specifically about API's anyone have any experience with that because we're, I mean we're generally just rolling out like a data governance effort at our company. And I'm in charge of my domain, but considering that we might start sharing data with others like citizen data scientists, not only traditional like reports in our system, but also guys. That's something that I never really touched before. Anyone has experience with best practices or suggestions, or any potential pitfalls that I'm not maybe looking at right now.
I mean, that's a loaded question because there's so much involved governance. If you like, I mean feel free to reach out to me and I could talk to you specifically about what we've done with several enterprise clients because we've set up, I personally set up. Well, personally had a team with setting up enterprise data catalogs, which includes obviously governance, but I'm I'll put my email in the chat and we could feel free to reach out to me afterwards.
Maybe not, not so much the logistics, but it's like, because I think we're concerned and then specifically I'm in HR so I have like HR data privacy demographics, a lot of a lot of sensitive data, and making that available through API sounds pretty exciting to a lot of citizen data scientists, but they're like yes we can finally get some data, do our own analyses. I want to support that but then I'm also like, that's a lot of data to put out there.
Yeah, we can do security so logistically I'm not really all that concerned but maybe more like expectation level or level setting or culture standpoint. Like how do you roll that out to people.
So, the last data set that I worked with with the APIs was around food safety data, which is super sensitive that's basically like HIPAA data for all of us. That's what it is for the agricultural world. And so they are very sensitive about who gets to see what and who gets access and pushing and pulling of sensitive information. So if you want we can chat about that if you're at Plastic Pumps and chat about it there, or we can connect on a zoom call or email whatever floats your boat.
Yeah, I think my thoughts were sort of regardless of the sort of like rollout method, the way you make that kind of data available, I think being sort of a facilitator or enabler kind of as long as it takes is really important so that you set up really clear expectations about like, what is this data good for? What does it support well versus what does it not support well? And that way you can sort of be riding along helping people think about like, in your case, compliance issues or those kinds of things in particular.
And it's a whole lot easier. And we found this over the course of trying to sort of course correct some big long kind of multi-year projects. It's a whole lot easier to do that at the beginning than later. So it's really exciting to roll stuff out. And also, it's really, really helpful and important to sort of like everybody look around the table and you go, OK, wait, let's do this slowly. Let's do this really carefully.
Oh, we still have two more, but it was eight. But we had three last weekend.
Thanks. That was shameless promo.
Mike's a rock star. Nobody knew that yet.
So is Ted, but he keeps it very quiet. Ted, we need some YouTube videos there, too.
