image: thumbnail.jpg
Data Science Hangout | Tanya Cashorali, TCB Analytics | Saving millions with a Shiny app
What was your biggest win for a client? (23:30)
We’ve had a lot of wins moving someone from Excel into a reproducible R pipeline and then displaying all the results in Shiny.
One example is a pharmaceutical company that was analyzing clinical trial data super manually, generating reports in Excel and putting together a PowerPoint weekly.
It probably took about 20 hours of multiple people’s time.
We made a Shiny app and that turned this into pretty much no time at all. They sent us the data and it was updated monthly.
They were able to then take that Shiny app to senior management and clinical trial managers to make decisions based on data very quickly, helping them understand which clinical trial sites were having problems.
Another similar one at another pharmaceutical company was focused on drug manufacturing, where there are a lot of things that can go wrong.
When there’s a contaminant in a batch, it can basically cause millions and millions of dollars in company loss because they have to shut down manufacturing completely until they identify the problem.
This involves going up and downstream of these different drug products to identify the issues. It was taking a team of 5-10 people sometimes 6 months to identify the problem. Meaning for 6 months, you’re not able to manufacture drugs.
We built a Shiny app that built out a D3 directed graph. This enabled one person to go and type in the drug compound and see everything up and downstream. It took that one person now maybe a week or several days to identify the problem now.
That’s an instance where you could say this is literally millions of savings from a Shiny app.
To me that just shows the power of R and the ability to really streamline manual processes.
Resources shared:
📕 Rachael shared Libby’s officerDemo: https://github.com/LibbyHeeren/OfficerDemo
📕 Daren shared the xaringan package for producing super impressive presentationshttps://github.com/yihui/xaringan
📕 Tanya shared Jesse’s Twitter as a great resource for learning Shiny: https://twitter.com/kierisi
📕 Javier shared xaringanthemer: https://pkg.garrickadenbuie.com/xaringanthemer/ and xaringanExtra: https://github.com/gadenbuie/xaringanExtra
📕 Tanya shared Hugo’s podcast: https://hugobowne.github.io/
📕 Tanya’s Twitter: https://twitter.com/tanyacash21
📕 Tanya’s presentation on “How to Hire and Test for Data Science Skills” https://tcbanalytics.com/2017/11/07/tanya-cashorali-presents-strata-2017-nyc-hire-test-data-skills/
Where to find more?
► Subscribe to Our Channel Here: https://bit.ly/2TzgcOu
► Data Science Hangout site: rstudio.com/data-science-hangout
► Add the Data Science Hangout to your calendar: rstd.io/datasciencehangout
Follow Us Here:
Website: https://www.rstudio.com
LinkedIn:https://www.linkedin.com/company/rstudio
Twitter: https://twitter.com/rstudio
Transcript#
This transcript was generated automatically and may contain errors.
Hi everybody, welcome to the Data Science Hangout. If you're joining for the first time, I'm Rachel, and it's great to meet you. If you've been here before, you've heard this feel a bunch of times before. But if this is your first Hangout, this is an open space for the whole data science community to connect and chat about data science leadership, questions you're facing, and what's really going on in the world of data science.
So the sessions are recorded and shared to the RStudio YouTube, as well as the Data Science Hangout site. So you can always go back and rewatch and find helpful resources too. I'll say it up front as well, because I always forget, we do have a LinkedIn group as well for the Hangout. So if you ever want to connect with people there or continue a certain discussion, you can use that too. I'll put it into the chat in just a moment.
We always want to create spaces where everybody can participate and we can hear from everyone. So there's three ways that you can ask questions today. All this is audience led. You can jump in by raising your hand on Zoom. You can put questions into the Zoom chat. And feel free to just put a little star next to it if you want me to read it out loud. Or I could just call on you to introduce yourself and add some context. And then also we have a Slido link where you can ask questions anonymously too. And I see Tyler just shared that in the chat. Thank you.
I like to reiterate, we love to hear from everyone, regardless of your level of experience or the area of work as well.
But with all that, today I'm so happy to be joined by my co-host. And I'll say, so excited that I even announced on Twitter a day ahead that I thought yesterday was Thursday. But so excited to be joined by my co-host for today, Tanya Casciarelli. Tanya is CEO and founder of TCV Analytics. And Tanya, I'd like to kick these off with just having you introduce yourself and sharing a little bit about the work that you do.
Sure. Yeah. I got nervous. I was like, oh no, is it Wednesday instead of Thursday? Yeah, it's been one of those weeks.
Tanya's background and TCB Analytics
I'm Tanya. I, let's say the short version, I got my career started, I dual majored in computer science and biology. So I started in the bioinformatics world. So probably unsurprisingly, that's how I got started using R. It was about 2005 or so, believe it or not. And I was working at a children's hospital informatics program for a bunch of really smart PhDs and had no clue what R is. He told me, go home this summer and read about R. So I actually printed like the entirety of the CRAN documentation and took it home and thought, oh man, this is like really, I thought it was just, I was in over my head, but yeah, I had some really great mentors and learned a lot.
And then I worked in biotech for a few years, a bunch of startups, and then I did some non healthcare stuff, a telecom startup, found myself back in healthcare at Biogen, which is a big, my first big company that I worked at and always knew I wanted to consult and kind of do stuff on the side. So I was doing that and only lasted there nine months, big company, just like wasn't for me, even though I learned a lot and had a great time, met a lot of cool people. And then I started TCB analytics in 2015 or so.
And we have clients kind of all industries, pharma, we do, we work with financial trading companies. We worked with the department of defense and we did some pretty cool AI machine learning stuff for the army. So really all over the map and just having fun because we get to kind of pick and choose our projects and I get to work with friends. So that was a short, short answer.
That's awesome. While we're waiting for questions to come in from everyone, I'm curious, what's something you're most excited about in data science in the next year or so?
Yeah, I, you know, I started, I was thinking about like what tools and stuff are exciting, but I don't even think it's tooling or anything like that. It's it's really that it's becoming more widespread and known and I feel like it's getting evangelized more and the more people I talk to people, a lot of you have probably had this experience before, like what the heck is data science? I have no idea what you're talking about. And usually I use the sports example because everyone kind of knows the movie money ball and fantasy sports. And I just think people are starting to get it more. And obviously that's good for us and helpful for us.
So it's simple answer, but yeah, I think I'm just excited that more and more people are just getting it. Like I thought years ago that CEOs at every company should be familiar enough with data and the basics of data analysis, the same way that they were expected to be familiar with email, like 30 years ago, everyone started using email. I think data is getting to that same point.
Like I thought years ago that CEOs at every company should be familiar enough with data and the basics of data analysis, the same way that they were expected to be familiar with email, like 30 years ago, everyone started using email. I think data is getting to that same point.
Big company vs. startup experience
Definitely. I see Libby, you asked the question in the chat, if you want to jump in.
Sure. I was wondering what about your experience made you decide that the big company was not the right one for you? Whether that was like wanting a bigger variety of projects or there were IT barriers or different stuff like that.
Yeah. I think the pace at which things moved, now at the time I was very burnt out from working at startups. So I kind of liked like breathing a little bit. And I think just a lot of politics to have to go through, trying to do seemingly simple tasks, it just was like a lot of bureaucracy. So you kind of felt like you had your hands tied in a lot of situations. That said, there were really good parts about it too, just learning how to navigate those big companies and the people that I met. But yeah, I'd say it was probably just, I felt like I couldn't really get much done or push the needle in certain ways that I could in a startup. Yeah. Lots of red tape, exactly.
Challenges in data analytics consulting
I see, Connor, you just asked a question as well, if you want to jump in next.
Hi, Tanya. I work here at RStudio with our consulting customers. So I would love to understand what you view as like the biggest challenges for data analytics consulting in 2022 or just in general.
Biggest challenges. It's always people. It's never R, it does everything. I can make R and Shiny do anything, but it comes down to, I think, I could tell when a client kind of knows exactly what they want. And most of our prospects and customers come to us saying, we know we need this. We love Shiny. We want to do things with Shiny and data. But I think the hardest part is getting people to understand why they're doing what they're doing. There's so many times where people just want to put all the data into a data lake or let's join all these tables and I'm like, why? So I'm always asking, well, what questions are you trying to answer? And more importantly, what action will be taken? And this sounds probably like a trope to everyone on this call, but it's amazing how many people and clients just, they want to check a box or they want to do the fun data thing, but they don't have any valuable business reasons for doing it. So that's a hard thing.
And then another thing, it's always scoping. I feel like with data products, it's much different from a web application because there's so many possibilities. As we're building things, we constantly run into situations and scenarios we could not have predicted until we start building it. So that's always tough, but we try to mitigate that by doing either retainer-based contracts or being really conservative with the fixed price and really scoping it out well.
Thank you. Yeah. And I noticed someone noticed my turntables. I used to DJ in college back when I was still cool in Boston, and I actually, I busted them back out during the pandemic. So that's what that is.
I think you may be the third data science leader slash DJ that we have. Really? Oh, who are the other? I got to know who the other ones are. Tanya, it's funny. This is Javier. There's another person here who lives in Southern California, Sep. I don't know if he's on this meetup today. He's involved with the SoCalR users group, but we've known each other for a few years and only this year realized we were both DJs. So we got together a few months ago and had a little garage DJ session. It was fun. Oh, that's awesome. Well, if you're down in Florida, you'll have to let me know.
Managing scope creep
Yeah. Tanya, I did have a quick question for you. You talked about making sure sort of your statement of works are very clear and concise and that, you know, you try and price project based. How do you, I don't think the problem is data science related necessarily, but how do you manage scope creep?
It's tough. Yeah. Cause I mean, there's, there's times where you give a little and, you know, you know, show of good faith where things are actually easy to do and you do it as like, all right, like we'll do this one thing, but then that can be a slippery slope and they think that now we can ask for all the things. So it comes down to, I think if you can itemize everything, you know, whether it's in GitHub issues or just JIRA anywhere, and then really prioritize that list. So I find it's really hard to do in a phase one because you're learning so much. So we try to, we try to estimate in like an exploratory sometimes one month long even to properly scope.
And that always, you know, it's never perfect, but I find for the second phase, and we just did this with a project. We know a lot more and they started asking for so many things. We kept a backlog and then we have our consultants go in and estimate. And we always pretty much, I wouldn't say double, but we add some padding onto that for uncertainties. And then you just throw it back to the client and say, look, you could purchase, you know, block of hours, 50 hours, a hundred hours, but here's the list. And now you put the onus on them to prioritize and say, okay, well, we definitely want this and this. And then even on top of that, I typically recommend another block of miscellaneous hours for like unknowns that come up. And that allows you to at least like quantify it because it becomes very like hand wavy if you don't have that. And if you can't back it into the money or the cost, it's tough. So I find fixed price, it can go either way. It can really bite you if it's not scoped well, or you can actually make some good margins on it. So we try to balance that with every project and I've just learned more and more over the years of doing it. It really just has taken a lot of trial and error and I've messed up for sure and learned lessons from it.
Understanding client needs
Thank you. I know. And one of the answers just before this, you said like the hardest thing is people and like having those conversations with people and they just want to put a bunch of data into a data lake. How do you approach those conversations?
Yeah. So one thing I like to do is, again, I always say, I try to say like what are people going to pay for? Like what are you building that your clients actually will put money towards? Number two is, again, what will they do with this information? Can they take action from it? An example I can think of is like when you log into E-Trade or Vanguard and you look at your stocks and it says that you've got a, you know, 100% gain on this stock, well, what is that? What do I want to do about that? I want to sell it. So typically there's a sell button like right there. You never have to like look and browse through five more pages to get to the sell button.
So I think that's a couple of questions I try to ask consistently is, you know, what will your clients pay for and what decisions are they trying to make? And then I'll do that same exercise with the people that are really like driving product management or the data analysis. So we'll actually sit with a mock-up, Balsamiq or, you know, name your favorite mock-up tool and we'll talk through it and we'll lay it out. And it's amazing the questions that come up and once people start thinking about it and seeing it and they think, oh yeah, actually we need a filter here because I didn't think that, you know, the date range would matter. We need a minimum threshold filter because we have so few data points in this state. But those are things that I learned as I build things in Shiny, to be honest. Just building it and showing it.
That's great. Darren, I see you have kind of a follow-up to this question as well. You want to jump in? Yeah, thanks, Rachel. Yeah, Javier, you kind of asked my question and I had a follow-up to that based on Tanya's response. So I've had a lot of experiences when working with other clients in previous jobs where they tell me I know what I want, I know what I don't want, but I don't know what I want. And so I'm kind of curious if you had those similar experiences and how you manage those expectations while also delivering a product that they don't really know what they want.
Yeah, that's like kind of where the art comes in, I think, is and also getting the right people involved on the client side. So if there's a subject matter expert, I always try to work with them. Or if there's someone that I know is a decision maker, because, yeah, it can be tough. Trying to think of another example. I think it's important not to also rush to building something. There's definitely rapid prototyping, but then there's like, well, let's actually take a minute and think. I try to list out all the most common questions and use cases and then also understand who are the end users. Because if this is a CMO using it versus a data analyst, one's going to be a very exploratory design or dashboard, and another one's going to be explanatory. Very simple, high-level KPIs, maybe some drill downs, right?
So yeah, it's tough. It really comes back to having those conversations kind of repeatedly over and over until it starts to click. But it's one of the hardest things we have to deal with. But then again, I get clients that come in and they know, and they start building it, and they're like, this is great. I hope that answers the question. But yeah, it's not an easy thing to navigate, for sure. That's why I like the Shiny prototyping, actually. Because you can kind of iterate until they say, OK, this is right, versus like a full software build that could take months.
And then you're kind of stuck with this big React thing and custom D3, and it's not the right thing.
See, Javier, you're commenting maybe to another question in the chat, too. But Mark, I see you have a question there as well. Do you want to jump in?
Yeah, sure. Can you guys hear me? Yep. OK. Sorry, my sound's been funky. Yeah, it was a follow-up to Darren's question.
And I'm thinking from my own experience. I don't have external clients. It's all internal for the company I work for. But have you ever had a scenario where up front your client is very clear about what it is that they want, the inputs and outputs, and it's very well-defined? And then you deliver that some time later, and then all of a sudden things have changed, and now they're wanting something else. And then if so, how do you deal with that scenario?
Yeah. And I see Javier's note said luckily over seven years, probably still only two or so also nightmare scenarios like that.
But it seemed like they were very clear up front on what they wanted. It turns out that they were not. And it's a tough situation. And also when you're dealing with people or contractors, things happen in life. And we've had on one occasion, a contractor just have to go pursue a full-time job in the middle of a project. So luckily, you have a bench to handle things like that.
But yeah, I'd say it really comes down to I keep revisiting the SOW and the scope and saying I literally will map it out. Here's what we said the deliverable was. Here's what we did. Here's a deliverable. Here's what we did. Here are some things we even did that were out of scope. So hopefully, you can get to a point where you can justify the end phase by using that SOW. Now, the problem is that we've had is it actually wasn't that clear. SOW is very high level and vague.
So then you run into a scenario where you're doing a whole bunch of extra work, and you're cutting into margins, and that sucks. So you try to wrap it up as quickly as possible with the minimum viable product or solution, and then go on to the next phase. And honestly, try to make up that cash in the next phase, because there's a lot of work that was done. And SOW, yeah, statement of work.
But yeah, to try to mitigate that, you can always, if they're happy with the end goal, regardless of how it went, they're going to pay you for more work. You can, I always try to make it up a little bit in the second phase by either over-delivering on things or just adding in a few more low-hanging fruit things we know we can do. But that is highly valuable. Like speed, some optimizations, let's parallelize this thing and give them a 10x factor speed boost. And that's hugely valuable for them, even if it doesn't take us a long time.
Because I always try to tell people, they're paying for your experience. Just because it takes you a couple of days, it would take their team maybe never, or a few weeks or months. So always value-based price as much as possible and not just your time. Because you've been learning, we've been learning this stuff and researching and doing it for years and years and years, right? That's what they're paying for.
Because I always try to tell people, they're paying for your experience. Just because it takes you a couple of days, it would take their team maybe never, or a few weeks or months. So always value-based price as much as possible and not just your time.
So that might be a long-winded question, or sorry, answer. The other thing I'll mention is, in the beginning, we realized one mistake was that the whole project was scoped by the data analysts, the users of the tool, and not the engineers. And it turned into much more of a software engineering problem. So there's a lot of complexity behind this tool that the analysts were using that they didn't see. It was like a magic black box to them. So next time, I'd probably make sure that I pull in an engineer and understand more of the technical requirements of the thing we were essentially rebuilding.
Yeah, that makes sense. Yeah, thanks. Sorry, Mark. Didn't mean to jump in on you.
Edgar, I see you also put a question into the chat if you want to jump in next. Oh, absolutely. And I apologize for the typo. But I've struggled over time to get a good sense of developing a burndown rate so that I know when to stop working on what I'm working because it's getting too big in size or way out of scope. Do you have any wisdom there to share?
Are you typically doing like hourly? Is it a fixed price? No, no. Well, I should kind of give a little background. I work for internal stakeholders, right? So it's more so if we think about having 2,000 hours available of people. And then at some point, you're trading off with other projects or other stakeholders because this one project, it's just turning into a Frankenstein or like, you know, how to kind of stop and say, you know, get back to the end of the line now. It turns up.
Yeah, it's tough. I, again, try to always get to an end phase that everyone's agreed upon. Like, OK, we're done with this phase. Let's stop here because we delivered X, Y, and Z, which we agreed upon. And now there's a bunch of extra things that are creeping in. And let's properly size those and estimate and start a new phase. And if you feel like, well, we need to charge more this time around, or we need to pull in different people, or sometimes, you know, if we know we don't need a senior statistician, for example, PhD, we could charge. We charge a lower hourly rate, of course, for like a data analyst. Or if there's no like, you know, crazy statistics or machine learning needed.
So that's one thing we've done to reduce cost is we realize, oh, we don't need like, you know, our highest rate card price. We can pull in someone else, but still, you know, still keep the charges the way it is. So, you know, but it all comes down, again, I think, to value base. Like, you know, if you can say, look, we've delivered, or if you can actually ask the client how much budget they have, that's always an ideal scenario. Because that means that they have budget, meaning it's important.
It also gives you an idea for what they value it to be. And we've heard budgets that are sometimes way under what we can accommodate. And we just say, you know, that's not for us. Sometimes they're much higher, and then we'll give them options. We'll give them tiers. Like, here's a bottom tier. It's very basic, middle, highest. And the highest we usually make kind of ridiculous and hope they don't pick it. Sometimes they do, right? They want all the bells and whistles. And we say, okay, we got to, like, size up the team a bit.
Got it. I see I can draw a parallel to what we still use, which is by when you need it done. Like, when do you present to your VP or your, you know, most senior leader? And that's, like, past that date, we don't continue working kind of thing. So just having a hard deadline where it's budgetary or time-driven. Thanks. Yep, no problem.
Sharing Shiny apps with clients
I see there's a few questions on Slido as well. And one was, how do you share your apps with your clients so they're hidden from your other clients? Do you have multiple servers?
I'm a huge fan of ShinyApps.io. So I have probably 10 client apps on ShinyApps. We've been totally good with just the, I think, $100 a month package that offers a certain number of hours in authentication. So I just, it's a matter of literally inviting someone via their work email address to the Shiny app. And we've had as much as, like, probably 30 people on ShinyApps and totally handled everything great for one app. And then another, some other client apps where there's only a few people, but they're power users and they're constantly using it. So ShinyApps has been, I haven't needed ArcConnect yet, but we do help clients that have their own ArcConnect internally. And then we just work with their internal ArcConnect team.
Biggest client wins
Nice, thank you. I see another anonymous question was, what was your biggest win for a client that was data science driven?
Win as in, like, I'm assuming, I mean, like, great, like a good success story. So that's a good question. I think that there's been a couple, one of them, it doesn't seem as, it doesn't sound super cool, but there was a small pharma company that was analyzing their clinical trial data and trying to understand what's called protocol deviations. So when a patient either disqualifies or doesn't get enrolled and something went wrong in the trial, whether it's like wrong inclusion criteria, like they had the wrong cholesterol measurement and they were accepted or something like that, they were doing all this super manually generating reports in Excel and putting together a PowerPoint weekly. And it probably took like 20 hours of multiple people's time.
We made a Shiny app and turned that into like pretty much no time at all. They sent us the data, it was updated monthly and they were able to take that to, you know, senior management and clinical trial managers and make decisions based off that data super quickly and understand which clinical trial sites were having problems, what were the biggest problems. And that is just, I think, huge versus like hand crafting all these charts over and over and over and wrangling the data. So we've had a lot of wins just from moving someone off Excel into a reproducible like R pipeline and then displaying all the results in Shiny.
Another similar one I wanted to mention is another pharmaceutical company. In drug manufacturing, there are a lot of things that can go wrong. And when there's a contaminant or something in a batch, it can basically, it's millions and millions of dollars of company loss because they shut down manufacturing completely until they identify the problem. And that involves going up and downstream of these different kind of drug products. And, you know, two of these molecules make something and then it goes into something else and something else. So it was taking a team of like five to 10 people, sometimes six months to identify the problem, meaning you're not manufacturing drugs at that time.
We built a Shiny app that built out a kind of D3 directed graph. So it enabled one person to go and type in the drug compound, see everything up and downstream. And it took one person now maybe a week or several days to identify that problem. So that's, you know, that's an instance where you could say this is literally millions of dollars of savings from a Shiny app, right? So that to me is cool. And that just shows the power of R and the ability to really streamline manual processes like that.
So that's, you know, that's an instance where you could say this is literally millions of dollars of savings from a Shiny app, right? So that to me is cool. And that just shows the power of R and the ability to really streamline manual processes like that.
It's an amazing example. I'm just writing that, writing that time stamp down. That's great.
Maintenance and project support
So I see there's a few others on Slido, but I see Niels, you asked a question in the chat too. Do you want to jump in? I can ask the question if you want to add context. How do you deal with maintenance of projects? How much is included or not included?
Yep, that's a good one too. Typically, we'll have a support rate depending on the person and the skills needed. And it's, we'll either do like a retainer, like you get 10 hours a week, five hours a week of support. And that's an average. And then they'll bill hourly and what they actually do. Or you just agree that it's going to be an average of certain number of hours per month. And some months is more or less, and you just, they just pay that fixed amount every month.
Typically, I think hourly works best for maintenance, just in case things go really haywire. But the biggest thing is agreeing on a reasonable SLA, so service level agreement. So do you need someone within like two hours? That's, we'll charge a lot more for that. Is it within 72 hours, three days, 48 hours? You know, that's, that's usually what we try to go for. Because I try to, I try to work with projects and clients where nothing is an actual emergency. Like I'm not a brain surgeon and, you know, data's fun, but there's usually nothing that's life-threatening. So making sure they have reasonable SLAs and not expected to like jump out of bed at 2 a.m. to fix something is good for your sanity.
Career progression in data science
Absolutely. Caitlin, I see you asked a great question in the chat too, if you want to jump in next. All right, so I was just wondering if you could touch a little bit on your progression through your data science roles. Like I was just looking at your LinkedIn and I saw you're an analyst, data scientist, CDO, VP, and I was just wondering how did you kind of know when it was your time to switch roles or?
Yeah, yeah, that's a great question. So my first job, I was learning a ton and just felt super way in over my head and I was like a bioinformatics scientist or something. And I just, I stayed there for about five years, got a few promotions to like start helping with the sales of what we were doing. So that was good, but eventually just wanted to try something else.
Got basically a similar role. It was kind of a sort of parallel move and I had a lot less responsibility, but it was a different, completely different industry. And I was learning instead of, I wasn't in bioinformatics, I'm more biotech. I was learning about distributed software and databases, but I felt like I wasn't being pushed and I didn't really, I wasn't learning much beyond about the distributed systems in terms of my data science skills. They were kind of stalling. Like I found myself supporting Salesforce and helping sales guys do very menial tasks which is great. Like I did get to, I started using Tableau there and I got more experience working with business leaders. And then I realized, okay, I want to try managing and having like a big role at a startup.
And so that's what I did next. I found a role at a telecom company and was really brought in to like transform their entire consulting business into a data products company. And this is a company that was operating very much as a consultancy, but they had really, really valuable data and a very good use case. They were helping telecom providers understand when and why people switched.
So back in the day, you know, you were always, you were going to T-Mobile, Verizon or Sprint and you didn't, everyone had a phone number eventually. And then it became a poaching game of like Verizon trying to steal T-Mobile's customers and vice versa. So they had all that data because you take your phone number with you when you switch. They had your zip code. They had, you know, a lot of different things.
So I saw that and I was like, well, that's valuable. And we, I could see why, you know, Verizon would want to know where high value T-Mobile customers are located in California. So anyway, that's a long way of saying I saw the value in the data. I felt like I could really make a difference there. People were like copy pasting SQL scripts and not saving anything they were doing and fat fingers in Excel and not able to reproduce past reports that we had sent to, you know, our biggest paying customer.
So yeah, I was brought into it and I helped hire people and I worked my ass off, but it was like a really good, great experience. And then I started to get a little burnt out. And I think that's when I decided to go back to healthcare and a big company. I kind of played the, I was a data scientist and just took my time to learn a bunch of things. And that's when I started consulting on the side. It became much too much. I was working at night and weekends and my full-time job. And I was like, it's time. I'm just going to talk to my boss.
Luckily, my boss was very understanding and he, we brought them on as a client. So now, you know, we were able to have Biogen as a client and these different companies that I worked for. So I think it's when you feel like your skills are stalling or you feel like you're not being challenged. I definitely felt like, okay, it's time to make a move and use everything I learned to like go make a difference somewhere.
That's really interesting. Thank you. Yeah. Thank you. That's really helpful.
Lisa, I know you put a question into the chat too, if you want to go next. Sure. Hopefully you can hear me okay.
So it, yeah, it looks like a lot of, or from what you described, a lot of your big wins are from, you know, maybe complex, maybe involving Shiny apps, data summaries and visualizations. And I'm just wondering, you know, how often you find yourself doing that type of work. And I'm really curious about if, you know, do your clients go in thinking like, oh, they're going to make us a complex data viz, or do they go in thinking we need AI and machine learning and you are kind of like, you know, pulling them back a little.
Yes, yes, yes. So like everything, basically I would, I always tell people, I would say we really only do machine learning, statistical modeling, maybe 20% of the time. A lot of work we do is, like you said, data summaries, visualization, automating manual processes and moving them out of Excel, a lot of Shiny app, Shiny development work.
I could give an example about the Department of Defense. They had budget for a big AI machine learning effort. We did actually end up doing some pretty cool machine learning for them. But like you said, a lot of companies think they need machine learning or AI, and they really don't, like they just need some summary statistics. And we can talk them off the ledge with that by usually just showing them what can be done. And then they see it and they're happy. They realize, okay, maybe that was, maybe that's down the road.
Department of Defense inventory project
But for the Department of Defense, they had an inventory problem. So every time anything is purchased in the army, anything from back to front, anything from batteries to laptops to Apache helicopter parts, it's all entered into a database. And it's entered pretty much manually on the field from a laptop. So there's a lot of typos and fat fingering, zeros that are O's, ones that are L's. And we're talking like 1.2, I want to say it was billion records or something. So there's no easy way to go in and just fuzzy match or deduplicate. So if the army doesn't know what's readily on hand in the inventory, how can they be ready for certain situations?
So we were brought in to do that, did some really cool, we used some genetic algorithms. I don't remember off the top of my head exactly which ones, but ways of comparing different fields and then quantifying things like the keyboard distance from different keys, common typos. And then using Levenstein distance for fuzzy matching and combining all these things into a very custom machine learning solution for them. And then using a Shiny app to expose the results. So we found a whole bunch of items that were clustered together because they were very similar, but they were duplicates. And so then we worked on scaling that with them and helping them find all those.
Now the AI part, we kept explaining them that's way off in the distance, right? You have to do the machine learning first, have a human in the loop. That's what the Shiny app did. It helped us then visualize the results and then agree with them and say, this is right, this is not. And that we kept explaining to them like that builds up your training set. So have someone come in, there is manual work involved. This isn't a magic button where a robot comes in and just does it all. So I think it's the hump between machine learning to AI that's really complicated because there's no way you turn on a switch and just let the machine automatically do this stuff, especially if it's very important, let's say healthcare related data.
So I think it's the hump between machine learning to AI that's really complicated because there's no way you turn on a switch and just let the machine automatically do this stuff, especially if it's very important, let's say healthcare related data.
So yeah, you're totally right. And we talked them off the ledge. And I wouldn't say that we've ever done artificial intelligence yet at this point. We've done a lot of machine learning, but getting to the AI part is a lot of, I think, human involved work to get there until you feel comfortable enough turning that switch and letting the machines take over.
Sensitive data and HIPAA compliance
Thank you. That's a really cool example. Let me see. There's a few other questions that were on Slido earlier I wanted to make sure I get to as well. And both were around dealing with sensitive data. Do you have experience working with clients who manage sensitive health data? Are there HIPAA compliant blueprints out there for hosting Shiny apps?
Oh, very good question. We have a lot. And we've struggled in the past with, you know, especially with, you know, especially with the drug manufacturing work we're doing. I had to do some research on making sure it was all, I can't remember the word now. It's not, well, FDA, you know, there's FDA regulations and there's all kinds of things you have to comply with. But for the most part, we haven't, you know, run into any issues where we just literally can't do something. People are typically trustworthy now or trusting of AWS and they have HIPAA compliant clusters. There's also a GovCloud. So we use that for the DoD. Luckily, the data we're working with, nothing was classified. So that helped. I didn't need like security clearance.
But yeah, it's, I think there's a lot of stuff out there now to support it. R can be validated in a clinical environment as well. And usually it involves just writing proper unit tests and making sure things are reproducible. So we've had the discussions, but it's never been like a showstopper. And usually it comes down to, well, we need everything in our cloud. So they have their own RConnect and it's behind their VPN and firewall. But yeah, the AWS HIPAA clusters we've used. I actually don't know though with Shiny Apps. I haven't run into that situation yet where it's been a problem, but I'd be curious. Like definitely feel free to follow up with me after too.
I think Rachel, you might share my Twitter or something at the end so people can reach out. Definitely. Yeah. And I think the other question touches upon this too, because it was asking around Shiny Apps IO. Isn't that hosted by RStudio? So what about the sensitive data? But I think you just answered as well.
Yeah. I typically just tell clients like, look, the AWS is going to be more secure probably than what you've got on site, but it really depends. So I think people are becoming more and more trusting of the cloud. So that helps.
Package management and Shiny deployment
Thank you. Javier, I see you put a question to the chat a bit ago as well, if you want to jump in. Yeah, thanks. It touches on ShinyApps.io as well, just about package management. Are you building project-specific packages or packages that contain modules or functions that you're reusing a lot? And if so, how are you deploying that to Shiny Apps?
Yeah. So actually, most recently, we started using Gollum, which makes production-grade Shiny applications and gives you a whole bunch of stuff like testing and documentation. But we didn't have a problem. We built our own package that did some custom D3 visualization for a pharma company. And we didn't deploy it on Shiny Apps, but they deployed it internally to our Connect server, and there were no problems there. So I haven't had any issues yet, or I haven't deployed my own package to Shiny Apps. But using Gollum, we had no problems deploying our own package to our Connect. So I don't know if that answers it, but... It does, yeah. That's awesome. Thanks.
Remote work and visiting client sites
Awesome. Tiffany, I see you just put a question in the chat, if you want to jump in. Hi. Yeah, I just was wondering if you do most of your work remotely, or if you ever get to visit the companies that you work with. And if you get to visit, do you find there's a lot of value seeing where that data is generated? For example, if you go visit a lab and you get to witness how the scientists are entering their data, maybe that gives you some greater insights on how to generate your apps to capture that and help them.
Yeah, yeah. That's an awesome question. Well, during a pandemic, of course, it was all remote. But yeah, we're a remote-based company. However, yeah, I used to go into client sites probably at least a few times a week. And oftentimes, yeah, I do think it's helpful. Not only can you understand how the data is being generated sometimes, but understanding the politics of situations too, where people work, how they work together, how often they're talking. There's just a lot of things you can pick up on when you are on site.
And of course, whiteboarding things is really nice. But one example I can think of that was pretty cool was we work for a health care devices company that creates the... They manufacture the rapid blood testing kits in hospitals. So if you're at a hospital, hopefully you're not, you get your blood drawn and they have the results very rapidly. That's the company that we do work for. And their problem was that they had all these faulty devices going out into the field at times, make things happen with hardware fails. But there's a ton of sensor data coming off of those machines.
And so we were brought in to do some anomaly detection, identify some spikes in particular proteins or problems before that hardware, that blood testing machine gets out into the fields at the various hospitals. And I got to tour that lab and it was a massive place with just all these machines. All these machines and they show me where the data's coming off of and how it gets put into a database. And I think just having that contextual knowledge and meeting with the people you're working with is invaluable.
And I was definitely a person that did not enjoy going to an office every single day, but I really loved being able to go to different client sites a few times a week. So I do miss it after this pandemic. Now I think I'll appreciate it even more, but yeah, it's super valuable. One of our clients is opening a drug manufacturing plant in Switzerland. So of course, I'm trying to go there.
That's great. I love how relaxed Porkchop is back there. He doesn't earn his keeper up here. He needs to work a little harder. He's our chief security officer.
Starting your own company
I was curious, Tanya, what has been the hardest thing about starting your own company? Oh, hardest thing. Well, there's been challenges. The first, when I gave my two weeks, there's obviously a huge panic. Like, wow, the checks are going to stop coming in. Yeah, get a job, Porkchop.
And there's that panic that sets in like, OK, this is really all up to me now. Now, if you have a spouse, of course, that helps. Like I had health insurance because I was able to go on my wife's insurance. And obviously, that helps. And I saved a bunch of money before making the leaps. I had an emergency fund and all that. But that's scary. I would say that's the hardest part is really just making the leap. But two weeks into it, I never looked back. I was like, this is it. This is awesome. It's working out.
And you just got to keep going. And it's nice to know it's all kind of on you. And you can't really blame other people now if things fail. Or there's always people involved. But it's nice to know that you're kind of in charge of if you want to make more money
one year, you work a lot more. If you don't care and you're OK, then that's fine. And I'm definitely a person, I'm enjoying work-life balance a lot more now. So I try not to overload myself.
But in consulting, it's also waves. It totally always happens that suddenly we have like 10 leads come in. And sometimes I have a hard time saying no because it's such an interesting project. And I want to work with these people. So I'll pull in more consultants to help if I need to.
So that stuff. And then also not having stuff like this, which is why I'm so glad that you do these Hangouts. Not interact with people every day in an office, obviously. Sometimes it's nice to just talk to other data nerds. So I have some friends that I talk to that also consult and will kind of commiserate about problems or ask for advice.
Building a community as a consultant
And the other thing I did was I started a Slack community of data scientists. So when I went out on my own, I knew I wanted to keep in touch with all these awesome people that I met throughout the years, whether at conferences or at jobs, college friends even. And so I invited them to a Slack community. It's kind of grown organically over the years. It's called Friendly Tech Space. And you can feel free to message me if you'd like an invite. But it's very much for networking. It's a private space. People talk about career development and our problems and Python. And then there's random channels too. So that's one thing I did to mitigate that problem as well. So it's kind of like keep my skills sharp and keep in touch with people.
Nice. Thank you. I was just looking in the Friendly Tech Space Slack too. Yeah, I haven't. Again, sorry. Are you having in-person events again with it? Not yet. No, we've done virtual tech talks. We have people sometimes give a talk. But we haven't done an in-person event since Boston and before the pandemic. And now that I'm in Florida, I don't know how many people are down here. But maybe we'll plan one for up north or something soon or try to get people at our studio conf together. But yeah, sometimes I'm not as active in there either because I'm just busy with client work. But it's awesome to have people just helping each other in there and talking about questions together.
Knowing your limits and finding your niche
Definitely. Libby, I see you had a question in the chat too. Yeah, she kind of actually answered it. Tanya, a little bit you were like, you mentioned that there were some projects where you were approached and you felt like it was too much and you needed to take on more people to help with it. Are there any people who have approached you with things that you felt like your team just couldn't solve? Like, this isn't worth it?
Yeah, yeah, for sure. I'd say we're gaining on the kind of DevOps, software, engineering, architecture side of things. But we've started to realize that's not really what we want to do or what we're good at. I mean, we're good at it enough, but it's not our sweet spot, right? It's not our bread and butter. We can do it. But number one, I prefer to do the analysis. I prefer to do the Shiny development and things like that. So we've accepted projects like that. We've done it and gotten through it and done a good job and delivered. But there's times where I just felt like we can do better work elsewhere.
So yeah, exactly. Figuring out your boundaries, not saying yes to everything. I've learned that over the years. In the beginning, you're hungry, right? You're saying yes to everything. You're like, I need to pay the bills and eat. And eventually, you can pick and choose things. And I think we're really starting to hone in now on a lot of Shiny work. We're really focusing there.
Where we excel is helping the business answer questions with data and tying that gap. There's just such a skills gap between people that can live in Excel and do so much stuff there. But then that lift going to something like R or Python is just, you're like a superstar. And you can do so much so fast. And people have asked me at jobs. They come over to my desk and say, how did you generate that plot so fast? I'm like, R and ggplot. And then they become a believer too.
So anyway, yeah, I think that we're starting to focus a lot more on what we're really good at. And it's been seven years. So it takes time.
Evangelizing R in an organization
I love those examples of things being so much faster with R and showing how you can do it. But I'm curious if there's someone who's listening and who's the only R user at their company. And they're trying to get people on board. That's been me. That's been me a lot of times.
Oh, I see Lisa's waving too. Yeah, so bless ggplot. I think Shiny is always the game changer. People see that. And they're like, ooh, you know.
I'd love to hear from someone else how you've evangelized R, what you've done being the solo R person.
No, go ahead, Lisa, please. I mean, I'm very new to my job. But I think there's still a sense that R is this statistical programming language, which it is. But I think the way to pull things in is actually with more like the parameterized reporting. And eventually stuff like Shiny, I need a lot more tooling set up to be able to do that nicely. But I don't think, at least not people in my organization, they don't think like, oh, R is great for making reproducible reports. And I'm like, no, it's really good at that.
So just showing off some of that stuff that people don't know about. Yeah, and I think we've taught a lot of R, too. And one thing that makes people kind of get it is we do this exercise where you give half the class a task to do in Excel, and half the class does it in R. And then we just change a parameter. We say, OK, now do the same thing, but for this baseball player instead of this one. And the R side is done in like two seconds. And they just change one thing at the top of the script. The Excel people are like going to get the data again, and having to wrangle it and do things. And it really gets people, I think, to like, oh, I get what scripting is now and why it's fast and reproducible.
And one thing that makes people kind of get it is we do this exercise where you give half the class a task to do in Excel, and half the class does it in R. And then we just change a parameter. We say, OK, now do the same thing, but for this baseball player instead of this one. And the R side is done in like two seconds. And it really gets people, I think, to like, oh, I get what scripting is now and why it's fast and reproducible.
That's such a good idea. Javier, I know you were about to jump in, too. No, I mean, Lisa kind of took the words out of my mouth. So same kind of thing. I partner a lot with, I mean, I used to be in finance. So I end up partnering often with our finance teams. And they generally see the value, let's say, right away because they understand the limitations of Excel or something like Excel that they might be using. And when they realize that you could very easily use tidier for pivoting long or pivoting wide, but instead of having like a million row cap or 5 million row cap, they could do this with a billion rows. They're always super impressed.
And then just like all the manual work that they have to do on a monthly basis for month end close or quarter end, seeing how you can automate a lot of that to just take out most of the manual work, I think impresses a lot. And then just like Shiny apps, like everyone's saying, or R&D presentations that are just knitted to HTML or websites, like jarring on websites. These are things that if you're comfortable with R&D, like jumping to those websites is pretty easy. And then seeing how you can automate all the elements within those websites also impresses people.
All right. Yeah. And one thing we were able to do in a Shiny app was, I was just going to say officer. Yeah. They were again, making manual PowerPoints and charts. And anytime I see, I see this in so many companies, they're just like copy, like Excel, copy, paste, and like, oh, it's just, and we streamlined it and we put it either in a Shiny app. So all the SQL is done for you. You pick a filter, whatever it is. Yeah. And officer lets you literally download a PowerPoint deck with SVG graphics, by the way. So you can edit the titles and the charts and everything. It's just that like people see and their mind is blown because it takes them, you know, days to put these PowerPoints together.
I was just sharing Libby's talk on officer. Oh, nice. Okay. Someone sent a question for me to ask, and it was, do you ever use plumber when you deliver a Shiny app to a customer? I have not actually, I know it's for APIs, but I have not used it yet. Probably on my list as soon as we have another use case for it though.
Resources and recommendations
Awesome. I want to make sure there weren't questions that I missed through Slido or through the chat earlier. So just want to remind everyone, if you want to raise your hand on Zoom, feel free to as well.
I was asking people this a lot in the beginning when we started these, but are there certain resources or podcasts or books that you'd recommend to us all?
Um, yeah, there's so much good stuff out there now. I don't know if, um, Jesse Mostapak is doing her live Twitch coding anymore, but she is on Twitter. I don't know if you can link her Twitter handle maybe, but she was like learning Shiny and in real time Twitch streaming it and people could come in and ask questions. And she's always a good resource because she was actually a high school teacher and turned nonprofit data scientist. So, um, she's just like a huge, she's like my resource when I need to go to, where do I find things to learn about? And then of course, you know, R4DS, uh, Hadley and, um, anything I've used RStudio documentation like beyond for Shiny, just, it's a great resource and there's good videos there.
Um, and then the, what's Hugo's podcast? Is he still doing that? That's usually a good one. I'm not sure what the name. Yeah, he started doing this one separate from, from Datacamp. I know that. We don't use Datacamp anymore. I can, um, I can look it up. Yeah, but the gradient. Oh, that's right. I have some of them in the chat too. And feel free if everyone else has others, feel free to share as well too. And stack overflow 60% of my code. Yes, I do a lot. There's a lot of that. Um, there's so many, I've been using R for what's 2017 years. Is that it? That's ridiculous. And I Google so many things, so I can't tell you how many times I copy paste how to rotate, um, axes and GG plot every time.
Imposter syndrome in data science
Okay. I see Libby, you asked another question too. And I'm really curious about this as well as I think about speaking at the RStudio conference. So do you want to ask that one?
Sure. Um, shameless plug for the podcast that I'm developing, which is kind of all about this is, um, imposter syndrome mitigation. I feel like there's a really big veil of mystery that's like drawn over data science jobs because we maybe don't speak to each other about what we do on a daily basis. Like, I don't know what all these other people do. They must be doing something that's so much more complex than what I know how to do. I must be terrible and not a very good data scientist. Um, so have you ever dealt with imposter syndrome and. You know, what did that feel like for you?
I, yeah, I think if anyone has an experience that they're doing something wrong, like, uh, cause I mean, there's so many talented people out there and like, you know, at my first job, especially, I, like I said, it was all like PhD level statisticians from Cornell. And, um, I just, you know, I'd be in three hour long conversations about them debating and use of a various certain prior for Bayesian algorithm. Anyway, I, um, yeah, of course. And I think, uh, what you have to realize too, is what people are putting out there and people on Twitter. That's, you know, oftentimes they're focused very much on one thing or it's their entire career and their job is based around it. And, um, I've learned, like, I don't have to be the best to any one thing, uh, but I can focus on what I am good at. And if that's, uh, helping, you know, with the technical communication between the business and the tech people, then I'll focus there. Like I'm, I don't have to be the best coder. I can bring in someone that I know is a D3 wizard, or I can bring in someone that I know is great at optimizing code. Um, I've always felt like I know enough to be dangerous. Um, but yeah, I think everyone deals with it. And at the end of the day, you know, there's way too much to know to absolutely feel like you're an expert at almost anything.
Um, I've always felt like I know enough to be dangerous. Um, but yeah, I think everyone deals with it. And at the end of the day, you know, there's way too much to know to absolutely feel like you're an expert at almost anything.
Like I, when I do time series forecasting for a client, it's, I use packages that are available. I, I try to, you know, consult with, if I need to consult with a statistician that I'm one of my good friends, I trust I will. Um, but there are people that spend their entire lives, like researching just one segment of time series forecasting. Right. So, um, yeah, just keep learning and it's possible to know it all. And, and remember the, keep the confidence that you should have, because if you're using R at all, you're already way ahead of, you know, most of the population.
Yes, job and job requirements are terrible. Actually, I have a talk about this entire topic and not to like shamelessly self plug, but I, I do kind of go off on a data science, uh, job descriptions. So I'll try to link that real quick. It was from Strata in New York, uh, a while back. Um, so I will link the talks, but yeah, I, I kind of, I go, there's all these crazy Venn diagrams of data science skills required. And, uh, it's all about that.
I know an hour goes by super quickly on me. So I just wanted to check if you have a hard stop or if you have a few more minutes, I have a few minutes. Okay, cool. Um, I see Parth, you asked the question in the chat.
Do you want to jump in live? Um, maybe you don't have a microphone. Um, yes, so I can ask it is shiny proxy, a good alternative to shiny server pro. Does anyone have experience with shiny proxy? That's the question.
I don't, so I can't speak to it either, but I, I remember using it like once way back when I was trying to get authentication for free and it was, uh, yeah, it ended up being, um, we just ended up going with a shiny server pro, I think. And now we just use shiny apps. Someone probably has used it though. And I know I've heard good things about it.
And Tiffany, I see you asked the question about conference. Do you want to jump in?
Yeah, I'm, um, so I'm currently a student trying to get my master's and I'm learning art in RStudio and kind of, I think at the beginning stages. And I saw that there was like some workshops at the conference offers, and I feel like the, in terms of the, the, the, the, the, the, the, the, the, the, the, the, the, the offers. And I feel like the intro one be matched to me.
Um, but I just didn't know if there was, it would be an overwhelming experience for someone in my position, or if I would have like some value from it.
I am super biased. So not the best person to answer this question, but would absolutely love for you to join us and would love to meet you there too, and get to all hang out in person, but yeah, I'll let other people answer too.
Definitely go to it. It is, I don't work for RStudio and it is the best conference I've ever been to. Like there is at least one thing every section session that I want to be at usually for. So, um, and as someone who's new, um, you will find something for you as well.
Yeah, I, I also definitely recommend going. It was the best conference I've also been to, um, and just the networking alone and the people you get to meet for sure is worth it. Like I, I met a few people that now I've worked with after, after the fact and learned a lot from absolutely.
It was the best conference I've also been to, um, and just the networking alone and the people you get to meet for sure is worth it.
And if anybody is going or anyone is thinking about going, and I know we all haven't been I know we all haven't been in person in a really long time. And so it might be like nerve wracking too. Um, so would love to like, I'll chat about it ahead of time. So we know we have a group of friends there too. So feel free to always reach out on LinkedIn too. And let me know.
And I see tattoos putting in some recommendations for the workshops too. There are a lot of awesome workshops as well.
Well, thank you so much, Tanya, for sharing your experience with us and just answering all our questions and thank you all for the questions too. That was fun. Yeah, no, this was great. Thank you for having me. The questions are awesome. And, um, I could talk about this stuff all day. So I didn't even know about this event for some reason. So I'm going to probably try to join on Thursdays now when I can. Awesome. We love to have you here. Cool. Bye everybody. Bye.