2020-05-29-lisa-shillerr.mp3

Lisa Shiller: [00:00:00] Whatever I'm doing for work or whatever I'm doing in my life, I am who I am. I got a very unique lifestyle. I work remotely from Mexico and most of times, sometimes I go to Toronto. And having the guts to just, like, be who I am and you know hide that when I'm working with a company, I think it's important to work with other people that are also who they are authentically. It's just like whoever you are at the end of the day, if you are not who you authentically are at work, that's what you spend most of your time. And I don't think that's a sustainable way of living and end an unhappy employee.

Harpreet Sahota: [00:00:55] What's up, everyone. Welcome to another episode of The Artists of Data Science. Be sure to follow the show on Instagram @theartistsofdatascience and on twitter @artistsofdata. I'll be sharing awesome tips and wisdom on Data science, as well as clips from the show. Join the Free Open Mastermind Slack channel by going to Bitly.com/artistsofdatascience. We'll keep you updated on bi weekly open office hours I'll be hosting for the community. I'm your host Harpreet Sahota. Let's ride this beat out into another awesome episode. And don't forget to subscribe. Great and review the show.

Harpreet Sahota: [00:01:44] Our guest today is a mathematician and Data scientist who loves dancing, cooking and adventure. She earned a bachelor's degree as well as a master's degree in mathematics, both from the University of Guelph, where her research focused on applying evolutionary algorithms to epidemiology problems. During her time at the University of Guelph, she published three articles, including an award winning paper on evolutionary computation. As experience Data Scientist, she's adept at collecting, analyzing and interpreting large datasets. She's most passionate about using her skills to make a positive impact, improve people's well-being, create sustainable abundance and decrease our carbon footprint by spreading awareness of sustainability. Throughout her career, she found interesting ways to combine her interests in health care and data science by working in organizations such as FIGGER 1; a case sharing network for health care professionals, where she founded a Data department. And she's currently at FoodMaestro, where she uses Data to tell stories, build recommendations and dive into products and trends to help her clients innovate and make smarter decisions. So please help me in welcoming our guest today. The founder of Toronto's Women's Data Group, a network for women that is shaping a Data culture in Toronto. Lisa Shiller.

Harpreet Sahota: [00:02:50] Lisa, thank you so much for taking time at your schedule to be here today. I really, really appreciate you being here.

Lisa Shiller: [00:02:55] Thank you for having me. It's an honor to be here.

Harpreet Sahota: [00:02:58] So talk to me a bit about your path into Data science. What sparked your interest? Where did you start? And how did you get to where you are today?

Lisa Shiller: [00:03:07] Well, my path into Data science really just comes from a love of mathematics which I've had since I was in grade six. I just really drove into math in my studies and becoming a Data scientist seemed like the natural progression for me because what do you do in industry when you have lots the Data, you apply lots of maths, so my journey just kind of flowed from passion and interests.

Harpreet Sahota: [00:03:34] I love that. I love that, because you kind of developed that passion through working on something for a long enough time that it just becomes kind of your craft, right?

Lisa Shiller: [00:03:43] Exactly. Yeah. So I mean, in university, the algorithms I was actually working with were Data science algorithms before there was actually a term data scientist. And so when I went into industry, I just naturally started applying mathematical algorithms to Data. And then I became a data scientist.

Harpreet Sahota: [00:04:04] Yeah, yeah. When I was coming up, it was just called statistics.

Harpreet Sahota: [00:04:08] So talk to us about the work you're doing at FoodMaestro. How are you applying data science to help deliver a better food experience?

Lisa Shiller: [00:04:17] So in FoodMaestro, we have pretty much all the data on food products in Canada and the UK. And that includes, you know, the ingredient lists and group includes other labels. Every text that you see on our product is Data that we have. What's layered on top of that are algorithms that can pass out, you know, these food products are for, you know, suitable for medium because they don't have any animal products in it or, you know, these other products are suitable for someone that has a tree nut allergy. And instead of having a customer read through all the ingredients, we have algorithms that just do all of that for you. So that's one aspect of what I'm doing at FoodMaestro. Another is creating products for category managers, which other people that work at big food retailers that decide what goes on the shelves. So how do we take sales data and product data to you know recommend using optimal products that should be having. So it's a lot of recommendations. The third area that I'm applying a lot of data science to is a personal interest of mine, is trading and sustainability algorithm. So for every product in the UK or almost every product in the UK, I have developed an algorithm that takes in all of this data from different sources on this products and says it gives a score for how sustainable it is. So then it will allow, you know, food retailers and shoppers to make smarter decisions around sustainability, which is super cool in my opinion.

Harpreet Sahota: [00:05:48] So that's really interesting that you say sustainability just for our audience out there, who's kind of familiar with that term or not familiar with what is meant by sustainability? Would you mind defining that for us?

Lisa Shiller: [00:06:01] For me and my work, what sustainability means is the impact on the environment as well as the impact on workers. Because if you think about the word sustain and also the impact on the farm and the ground farming practices, whether it's organic or GMO, one piece that I think people are about is, is the workers. I think that Fair trade is really important because the word sustainability is like, if you think about it, it's like sustained practices, like practices that will could keep going and going and going without feeding. And if you think about that which food or your your action and how you live, you know, it doesn't just impacts the environment. How are the worker systems like are they being paid an hour prior so that they can survive and be sustained as well. It's the C02 content of water usage. It's the farming practices of like organic or GMO. It's the packaging as well. Is packaging made from recycled material? Or  does it have the ability to be recycled easily? So it's all of those things. 

Harpreet Sahota: [00:07:15] That's really interesting. So how do you think Data science will impact clinical health, wellness, and sustainability even in the next two to five years?

Lisa Shiller: [00:07:26] Well, we have tons of data in the world. It's about making informed decisions. If I take the example of food because it's relevant and top of mind, if I care about sustainability and I walk into a store right now, there's no one for like I don't know what is sustainable and what not. But if that information was served to me and I was able to actually see through like someone interpreting the data that this product is sustainable and healthy, then I'll be able to make a better decision. And so when it comes to that, like what if you have the information presented to you super easily and then you just be able to make a decision because you have the information and something that you care about. And that goes for sustainability and health and not just in food. But, you know, let's say I mean, one of the industries is doing a really good job right now is that, you know, exercise. So we have, like, all the technology that, you know, if I care about exercise, there is this technology that has my Data and we'll be able to say, like, you know, you haven't taken enough steps today. So, you know, if you care about your health, like, maybe you should do more exercise or something like that, it's all about taking the data that we have, interpreting it and allowing just like everyday people to have access to information to make smarter, healthier decisions.

Harpreet Sahota: [00:08:48] In what ways do you feel we can leverage data science to help reduce our carbon footprint and promote sustainability?

Lisa Shiller: [00:08:55] Again, I'm going to relate this back to food. That's the area that I have a lot of experience in. Reducing our carbon footprint when it comes to food has to do with Data. If I want to reduce my carbon footprint or maybe even if that's not top of mind, just having that information presented to me, like, you know, when I go in and buy food or make any purchase, what is a carbon impact of that? And then if you want to generalize, it's how people live their everyday lives. A lot of it has to do. It's like buying local. It didn't have to travel very far to get to you. It's just about designing your life so that you don't have to commute super far, I mean, so much of our CO2, or has to do with transportation. So if you think about how transportation impacts every part of your lives and then you just reduce that, you're making such a big difference already.

Harpreet Sahota: [00:09:45] In what ways do you think Data science will have a big impact or at least the biggest positive impact on people's food choices in the next two to five years?

Lisa Shiller: [00:09:54] I think that has to do with presenting the information to customers in an effective way. And it's hard to do that in person, like in a store of brick and mortar store. You can't have like I mean, I guess you could, but, you know, having a little label like this is the environmental impact of this food item on the shelf. I just don't see how that would work. But shopping online, that  information is so easily presentable and having like a Data science or just, you know, some kind of algorithm in the background, working, collecting data and showing that's, you know, a number of like something to say how sustainable a product is, would make online purchases a lot more sustainable. And I really think that, you know, now it's like corona virus, most people are shopping online, and I think that we've kind of passed this threshold of online shopping that, you know, the convenience of it and just how many people have adopted that behavior. I personally think it's going to stick around. So there'll be more people shopping online and then they'll have more algorithms powering the information behind the products on these websites or apps or whatever people are shopping.

Harpreet Sahota: [00:11:07] Truly interesting, like a sustainability score on any product that you purchase that kind of comes up like. I don't say like a warning label, but almost like kind of how we have a food packaging labels that you mentioned like some kind of sustainability score on that package. Would there have to be some type of governing body that that regulates that if we were to go in that direction?

Lisa Shiller: [00:11:29] Yeah. So right now what I'm building is being verified by organizations. I think it would be a fantastic to get to a point where maybe it's like a, you know, like causers the FDA. It's like also, you know, there's some kind of governmental body that verifies the data behind the algorithms and the algorithm itself is powering some kind of a quote unquote, sustainability. You know, maybe we'll get to that point in a few years. That would be fantastic, because that also means that the government is just one step further to, you know, being concerned about the environment and and sharing that concern with the population.

Harpreet Sahota: [00:12:06] Talk to us about the project you worked on, where you use math data science to predict COVID-19 in the state of Guanajuato, Mexico.

Lisa Shiller: [00:12:13] Yeah. So this was a project I was just sitting in my apartment in Mexico. And it just being very weird that the information was shared with the population of Mexico is very little. It's very inaccurate. There's all kinds of politics about what information can be shared and what numbers are good and what numbers are bad, there's now so much of the Mexican culture is in grained with tourism. And so it was kind of a scary situation for Mexico. I understand why that information is mostly hidden or was mostly hidden. And I've done a lot of research on epidemiology, which is the study of how disease spreads through populations. That's, you know, all of my published papers are on and I spent several years of my life just researching this, sitting on the couch and playing with some numbers, trying to figure out when this thing was going to hit Mexico hard. And, you know, when it would be over. And what are the different, like strategies that we can take to mitigate And just playing with those numbers and my neighbor walked in and I was like, hey, like this is really cool. I think maybe you should share this information with people because we all really want to know. So I created an algorithm that models how COVID-19, was going to spread through the population here in the state of Guanajuanto. And I published it and it just kind of flew up this information that no one really had. And I don't think there's a lot of epidemiologists in that state. So there's nobody that was kind of taking on the modeling, the spread. So that's kind of where it came from and I had a really, really good impact on the side here. Yeah, I was like these papers and governmental bodies and everyone just like really needing this information.

Harpreet Sahota: [00:14:09] The blog post I read you talked about the, it's called the SEIR model, could you break that down for us. What is that and how did you kind of use that in this use case.

Lisa Shiller: [00:14:21] Yeah. So in SEIR model. It stands for Susceptible, Expose, Infected and Recovered. And this is a very popular type of epidemiological model where you have a population of people and at any given time you're in one of these four states, when it comes to coronavirus most people start in the susceptible phase. And then as time progresses, you move in to the exposed phase, which means you have contracted the virus. But you are not showing symptoms yet. And you're not able to infect other people yet. And then the infectious stage is where you're able to infect other people and spread the disease. And then recovered is actually you've the disease, you either recover with immunity, meaning you cannot get the disease again or you die. But either way, it is the same. And, you know, there's other setups on this model like we have SIR such as Susceptible, Infected, Recover and have the exposed phase. There is SIRS where like recovered doesn't mean that you have immunity if you move into the susceptible phase again. That's a model that.

Harpreet Sahota: [00:15:37] It's really interesting. And I think that kind of speaks to you know the importance of, first of all, understanding that domain. Like, I think a lot of people are just creating these really cool tableau dashboards, these really cool visualizations or whatever in Python and posting them. I see them all over LinkedIn. But without having a real understanding of the mechanism for which a disease progresses through a population, that it's kind of irresponsible to create these these type of dashboards, these models if you don't understand that kind of underlying mechanism. So thank you for for talking about that. And, you know, really giving the audience something to go and research on their own so that they become more knowledgeable about these other models. Can you talk to us about some of the key assumptions that you made and maybe even touch on the importance of having good or strong assumptions when you're undertaking a new project?

Lisa Shiller: [00:16:29] You know, mathematical models are only as accurate as their inputs. So everything relies on on what data you're feeding it and what the assumptions are and if you treat those assumptions even a little bit, sometimes that can completely change the results of the model. You know, that's one of the things that I contrast in. And sharing this information with people was the Data is not totally accurate. Did my best to weed through it and figure out what the truth is. But I did edit the article and model based on new information. But some of the assumptions that I had were, you know, the starting first of all, like. Yeah, like the starting Data. So how many people were actually infected, were actually already recovered and so on. So those are some of the inputs that you just have to figure out as best as you can. There's an infectious rate, which is the biggest one, because this fluctuates even a little bit. And it just seems as completely. It was deduced from a study that I found that I think was a given a scenario. It's the closest thing to a scientific study that we had. What I found was that infectious rate was 2.28 million for every person that's infected, they will pass it along to 2.28 people on average. And I played around at this number and said, OK, if we go into full lock down, this is how that number will change and lower and so on. And I tried a bunch of different scenarios and those in the article. And there's a asymptomatic rate, which is 17% that I assumed in the article, 17.9% and this was taken from the Diamond Princess cruise ship, which is kind of a scenario that's the closest thing to a scientific experiments just because it was an enclosed environment and everyone that left the ship was followed and kept in isolation. So we have very good numbers on  based on that study. Oh, yeah. The death rate and a mortality rate, which only really matters for just if you want to split up the recovered into fully actually recovered immunity or recovered meaning like they will no longer get the disease because they unfortunately ended up as a charity. So is that and then, you know, it went a little bit more into how many hospital beds I think would be in Guanajuato, which I have no idea. But that's just a guess. But I share that information.

Harpreet Sahota: [00:19:15] What's up, artists? Be sure to join the free, open, Mastermind slack community. Going to bitly.com/artistsofdatascience. It's a great environment for us to talk all things Data science, to learn together, to grow together. And I'll also keep you updated on the open biweekly office hours I'll be hosting for our community. Check out the show on Instagram @theartistsofdatascience. Follow us on Twitter at @ArtistsOfData. Look forward to seeing you all there.

Harpreet Sahota: [00:19:44] Can you maybe speak to what you would say was the most interesting or important finding that you got from this project?

Lisa Shiller: [00:19:52] I have two that I'll share, one is just in terms of the Data and the results. One of the things that I that I tested, which I think is very controversial and I don't actually know if it would work. So I want to preface it with that. But it's the best outcome came in terms of economics, because obviously, if we want to lock down, the virus would be gone very quickly. Who knows what happened when you get it up again. But that is not a good economic situation for Mexico because the economics are very different than Canada states. So taking economics into account, the best scenario is actually isolating the people that were vulnerable and at high risk of having severe case, isolating those people and then just letting the rest of the population go through life like relatively normally. The idea is so that the normal population would gain herd immunity at that point, it will be safe for everyone else to kind of join back into society because that herd immunity will have been reached. But we don't know if this would actually work in real life for those very interesting symbolic. The other is just a reaction from the community, which I was not expecting. I thought maybe 10 friends would read this article and they'd get, you know, some good findings from it and it would help them and either ease their fear or kick them into gear. So the reaction with was much more than that, at least in thousands of comments on Facebook threads and on my Web site. Lots of really great discussion and governmental bodies reaching out saying, hey, can you help us further with this analysis? And also just individuals sharing our gratitude. It was like so clear that people need information. People need to have the data interpreted and presented to them in a way that they understand it. And I was not expecting that. So it's very interesting.

Harpreet Sahota: [00:21:54] One last definition for us here. Can you define what herd immunity is?

Lisa Shiller: [00:22:01] Yeah. So herd immunity is when a certain percentage - and it's different for every disease, every infectious disease - when there's a certain percentage of people in a population that are immune to the disease, meaning they can no longer get it. And that immunity comes through either a vaccine or a previous infection when there's a certain percentage of people that can no longer get the disease. The trajectory of the disease is cut off. It actually stops spreading. It's unclear what percentage of the population would need to have immunity over COVID 19. It's estimated to be about 67%. But it's a range and we won't really know until we have a better understanding of the disease. That's how vaccines work. Not every person in the population needs to be vaccinated in order to eradicate the disease. We Just need to get the percentage of people.

Harpreet Sahota: [00:22:54] Thank you so much for that. So kind of shifting gears here now with our questions. You've talk about your project. You know, definitely very scientific in the way you went about your methodology and everything. But I'm curious to see, how do you view data science? Do you view it as an art or as a science?

Lisa Shiller: [00:23:13] I view it as the dance between both because it's a tricky thing. You have to have a creative mind. When I was studying mathematics very intensely. One of the things that I realized quickly was how creative mathematics is. Everyone has this idea that it's just, you know, logical like this means that this will always mean that and so on. But when it comes to proving things or designing, analysis, it's a totally creative process that you need to have creativity to do that. And that's why it's very creative. And then, you know, also scientifically, because there are certain rules ATS follow in order to prove or present information in a truthful way. So definitely the dance between both. And I think that most scientific you know, even if you look at civics, it's also very creative. But you have to, like, follow rules.

Harpreet Sahota: [00:24:08] So let's talk about how the creative process manifests itself in mathematics and Data science. How do you see it manifesting itself?

Lisa Shiller: [00:24:18] So I'll take maybe some real experiences, and example. So, you know, I'm presented with a data set, I need to, let's say, create a recommendation algorithm based on that data set. The first is just getting familiar with the data. Looking at it, being curious about it, trying different you know, visualizations of the Data, one of the things I love doing is I love working with Tableau because I just plug a Data set in there, play around with the different fields and just get an idea of the landscape of the Data is. And then in that process, it's you know, I'm discovering like how I could use this Data in order to make recommendations, all different types and trying different things. And so it's really not the discovery phase with all the creativity happen. And then once a few light bulbs go off, it's like, OK, now I will use the rules and the scientific aspects of what I know of their recommendation algorithms. And I will take that creativity and sheet it into those rules and then create the recommendation algorithm.

Harpreet Sahota: [00:25:28] You've got some awesome experience building a Data science practice from the ground up. What do you think are the essentials to lay the foundation on which the House of Data can be built?

Lisa Shiller: [00:25:40] So I'm a data scientist at a company and I'm trying to lay the foundation of Data science there?

Harpreet Sahota: [00:25:46] Yeah, yeah.

Lisa Shiller: [00:25:47] Well, the first thing that I think is really important is actually getting to bear with the way things are set up right now. And coming up with a plan about how to improve that. And usually, you know, companies can be especially startups and be at any level of like having their Data in Excel spreadsheets or having it a SQL database. But I think getting the Data to a point where, you know, it's in a database, it can be easily like analyzing and plugging in SQL and Python further important that having access to data to both of those languages, now it's like a technical infrastructure part. The second is creating a Data science culture. Data scientists have this knowledge of how algorithms can help. Nobody else in the company usually has that knowledge. So it's important to like create a culture where people have a general understanding of what Data science is and how it can help their projects or spark new projects. Having that basic culture is really important. Also going around and looking at every project and having a great understanding of what's going on and then what is important to the business so that, you know, you can kind of, as a data scientist recommends like, oh, you're doing it this way. If you actually have an algorithm at this point in this process, it'll make us a lot more accurate or it'll make it a lot faster or whatever. So that's something that I think is really important, like for creating a data science practice. And then actually coming up with plans with the different projects, like a data scientist is almost this internal consultant. For this project needs a data scientist role a little bit and there will be graft in there, another project and so on. And then also having the guts to recommend new projects where you see opportunities of how can you know, Data science can improve the process and make things better and then just actually having the guts to recommended that, you know, maybe there's new projects that there.

Harpreet Sahota: [00:28:02] So for someone who's the first data scientist in an organization and who wants to kind of cultivate that Data science culture, what are some challenges you foresee them facing and maybe share some tips on how we can overcome those challenges?

Lisa Shiller: [00:28:19] Yeah. So any data scientist, first data scientists at a company like most of the challenge comes from the culture. I've had pushed back before where people were doing things one way. And, you know, I'm not saying that their way is wrong and it's just I have skills, I could make that process a lot easier and a lot faster. But it's difficult to go in and say, I have all the answers. What I am doing is right. What you're doing is wrong. If you frame it like that, it's not going to go over well. So it's about culture and also just about approaching things from a perspective of like education. So it's like I'm educating you on these skills that I have and I would really like to help you. Another is, a lot of the time companies will be like, we need a data scientist because our investors say that we need a data scientist, I have no idea what a data scientist does, but we need you and I can present a lot of issues because from my experience, those companies don't necessarily actually need to do the scientists at that point. So I kind of just sit there twiddling their thumbs for a little while before you can come up with how we can help. So I think a lot of the initial conversation should be about, OK, what are the actual needs of the company and how can I actually help or kind of taking on big projects like joining a new company and not knowing exactly what to do.

Harpreet Sahota: [00:29:45] So for those organizations out there that are, you know, looking to to bring on Data scientists for the first time or maybe people out there who are looking to add Data scientists to their own teams. What do you look for? Like, you know, what is it that you look for in a Data science candidate? And do you know, do you have any tips on how someone can cultivate those qualities for themselves?

Lisa Shiller: [00:30:07] I think there are so many potential Data scientists out there, lots of really smart people. So when it comes to the technical side that are almost less of a concern to me, because if you have someone that's really smart, you can teach them. You can teach them Python or you can teach them SQL. It's really about creative side that I was talking about, regional look at a problem and creatively solve it, coming up with solutions that are you know, they don't show in a square box. I think you have that ability in your brain to think outside of the box. That's the most important thing. And the other is on a personal note, I think, you know, whatever I'm doing for work, or whatever I'm doing in my life, I am who I am, I have a very unique lifestyle. I work remotely from Mexico. But sometimes I go to Toronto. And having the guts to just, like, be who I am and not kind of, you know, hide that when I'm working with a company. And, you know, if I want to show up to work wearing like neon bright pink t shirt, like, that's just who I am. And I think it's important to, you know, work with other people that are also who they are authentically. And that's not to say that, like, everyone's going to want to wear bright pink t shirts, but it's just like whoever you are, I try to find that authentic person. And when looking at candidates, because at the end of the day, if you are not who you authentically are at work, that's where you spend most of your time. And I don't think that's a sustainable way of living and it's going to end up in an unhappy employee.

Harpreet Sahota: [00:31:56] I absolutely love that. Just don't live up to or don't try to be what you think a Data scientist should be like. Just be yourself as a data scientist.

Lisa Shiller: [00:32:06] Exactly. And you're going to be so much happier. 

Harpreet Sahota: [00:32:14]  So we talked a little bit about some of these you mentioned, you know, non-technical skills that are really important. But I think a lot of up and coming Data scientists, they tend to focus primarily on these hard technical skills and they think that that's really what's going to separate them from the rest of crowd. And, you know, you've touched on this a little bit already. So sorry if it's said to be repetitive here, but what are some of these soft skills that candidates are missing that are really in a separate from their competition?

Lisa Shiller: [00:32:41] Yeah, for sure. One of the most important soft skills is the ability to take a complex start or process and explain that in a very simplified way that anybody can understand. As a data scientist, you have all this technical knowledge, everything, you know, complicated in your brain and convoluted, and you're solving these problems with the Data. And then how do you actually explain that to the CEO of the company or the project manager. These people usually don't have the same technical knowledge that you have, and that's why they hired you. So it's about being able to explain that to those people. And also, a lot of the time, I work with external clients. I'm, you know, explaining how I came up with an algorithm that recommends products. I'm not going to tell them how the entire algorithm works. I'm just going to have to know what she says will matter to them for them to wrap their brain around what's happening. So that's a really important skills, data visualization skills. I think it's kind of between the technical and non-technical skills because you have to have the creativity and the eye to make a visualization that explains the Data and the best way possible, the ability to. And this is a tricky one, the ability to work alone and with a team, because I think this goes for a lot of engineers as well. But a lot of what I do is me just sitting at my computer and focusing very intensely by myself. And then a lot of it is also working with a team about the results on creating and, you know, the needs of the companies being able to flow between those two. You know, usually with people one or the other comes naturally for Data scientist, I think you've got to have both.

Harpreet Sahota: [00:34:30] So do you have any tips for for Data scientists who maybe find themselves in a room full of executives or maybe stakeholders that aren't so technical or external clients, do you have any tips for them on how they can clearly communicate their ideas, their technical concepts without completely losing their audience?

Lisa Shiller: [00:34:48] I don't want to make it sound like I'm calling the executives dumb because that's not what I'm doing. But what I'm recommending is talk to them like you're talking to two year old. Maybe not two year old, maybe like an eight year old. And then that's because you have to assume that you don't have any of that technical knowledge. So just pretend you're talking like your eight year old niece and you're explaining these concepts. And I think that's how you would get information across very effectively. Also, just have that perspective of they're looking to you for answers. They're looking to you. You're there because you have skills that they don't have. Most of the time just haven't acknowledged there have the confidence.

Harpreet Sahota: [00:35:32] So we talked a little bit about working on a team in a team environment. I'm just curious like if you have any tips for Data scientists who are in a team environment, but they might be scared of looking like they don't know something. But they do want to openly communicate that with the team.

Lisa Shiller: [00:35:52] I use to struggle with this one a lot because when I was working with companies, I would be the expert and I don't know every single algorithm. Googling things kind of, you know, a lot at a time where I'm just like okay, I think I know the generally the idea of how to solve this problem. But like, I need an algorithm that does this and this and that and then Google and figure it out and maybe look on a bunch of forums to see who has done it before. And that's OK. It's like you are not. It is just knowing that you're not expected to know every single thing. But you have the skills to be able to interpret and learn and communicating that also. Like, one day I just had this epiphany where I'm like am I not expected to know all these things and that's say okay, I can share that. And the people that I work with know that now. And I'll say I know how generally I want to solve this problem. I don't know what algorithm I want to use yet. I don't know exactly how it's in my look, but I'll figure it out. And just framing things like that where it's like, I don't know everything right now, but I will figure it out. And that's totally OK. I don't know. I used to think that's  that wasn't okay, but it's okay.

Harpreet Sahota: [00:37:07] I absolutely love that. And I think that's probably the most liberating epiphany I've ever had, is the fact that actually, hey, nobody really expects me to know everything. And I don't have to pretend like I know everything. It's OK to be like, you know what? I don't know right now, but I can go find that out. I've got you know, I've got the resourcefulness to be able to go figure this out. Just give me a day or two. Once I realize that it's OK not to know everything, like life became so much better.

Lisa Shiller: [00:37:34] Yeah and that applies to everything and not just data science

Harpreet Sahota: [00:37:39] Growth mindset. I'm not sure if you're familiar with Carol Dueck's work on the growth mindset?

Lisa Shiller: [00:37:43] No. But it sounds like I should be interested.

Harpreet Sahota: [00:37:48] Yeah. Definitely check it out. I think you'll like it, but it's pretty much just, if you distill it down, this is the belief that you're on long enough time line, you can figure anything out and you just have to have that mentality in that point of view.

Harpreet Sahota: [00:38:08] Are you an aspiring Data scientist struggling to break into the field, you can check-Out dsdj.co/artists, to reserve your spot for a free informational webinar on how you can break into the field. It's going to be filled with amazing tips that are specifically designed to help you land your first job. Check it out dsdj.co/artists.

Harpreet Sahota: [00:38:33] So I wanted to jump in to talking about your experience being a woman in tech. And if you have any advice or words of encouragement for our women in the audience who are breaking in to or currently in tech.

Lisa Shiller: [00:38:51] Yeah, definitely. So being a women in tech means that you will most likely be one of the only women in the room. A lot of the time, unfortunately. And I think that's changing. There is definitely more women entering tech, which I am so happy about because, you know, I don't know. It's just it doesn't make sense that it's mostly men. I think what that means right now being usually one of the only women in the room is that there is a subconscious thing that happens if you're at all familiar with imposter syndrome, you'll know what I'm about to say. That there's a subconscious. I am different than them. And they all think one way and I think a different way. And it's just it's so subconscious. I know before I became aware of imposter syndrome, I kind of had a really shy voice in my field. And once I started to research imposter syndrome and what that means, it's just almost being aware of it, eliminated it. I think it's really important to be aware of it and do it every you need to do to know that you are good enough. You are good enough. There's nothing wrong with you. It's nothing wrong to anyone else. You're good enough. And if you can almost work to your advantage as well, because, you know, once you establish that you are different from everyone else and that's OK. Then your creative ideas and the ways, the creative ways that you solve problems, you're going to have a bigger voice for that and you're not going to be afraid to share those ideas. So I think just be aware and be gentle with yourself and know that you just want to get sufficient, you know, there's nothing wrong with you and you're good enough.

Harpreet Sahota: [00:40:44] Can you talk to us about how you kind of grappled with imposter syndrome and how you overcame that?

Lisa Shiller: [00:40:49] It really was community. And that's why I created a Toronto Women's Data group. When I started working for this company called  Figure-One and there were those employees that was I guess we need a data scientist, but we don't know the things a data scientist does. So they actually sent me  to San Francisco. I went to a conference there and they were also just say, why did you meet with a bunch of these different companies? Like, you know, I met with Uber and Lyft. I met with, like, another healthcare startup. And I met with all the scientists there to try to figure out how I wanted to set up data science that fit your line. And in that process, I ended up, I didn't meet up that with all women in Data science. And it was women from like Netflix and all these other, you know, to new core and tech startups. I know that exist in San Francisco. And it was just like all women, it's all super smart and just helping each other out. And that was this moment where I was like, oh, my God, I've had been, you know, thinking that I'm not good enough up until this point where I'm seeing all these really powerful women and they're just like in community together, what was created there was a space where, you know, women can just be themselves, which I think also comes as imposter syndrome. I was feeling like you can't really be yourself. There's all these women that were just being themselves and helping each other and Problem-Solving and then, you know, I'm assuming would go back to the companies and will be like oh  I actually have this new perspective on this problem and [inaudible] and create and then I came back to Toronto and created the Toronto Women's Data Group. And in that sense, it allowed me to get over my own imposter syndrome because I had now, like, notice that it was there and then created my own community, like literally giving away our community. So I had a bunch of women that were super powerful and super smart and we're all helping each other out and teaching each other. You know, I think that I killed my own imposter syndrome and I didn't get hold of a lot of other peoples' as well and continues to do so even though I'm not really part of the organization anymore.

Harpreet Sahota: [00:43:03] So what can the Data community as a whole do to foster inclusion of women in Data science and AI? And maybe what can men do to help foster the inclusion of women in Data science and AI?

Lisa Shiller: [00:43:16] And I think it's acknowledging that, that sometimes it can be difficult to be the only woman in the room. And for reasons that I think most men don't understand. But just knowing that a imposter syndrome is a thing and creating space for opinions, women like not just going around the table asking others not even asking, but just like, you know, like giving time for women to state their opinions, asking them what their opinions are. Rather than assuming that they will interject when they have opinion. And once you do that enough times, it will feel to that woman like, oh, my opinion does actually matter here, even though subconsciously I thought it didn't, but it actually does so it's great. And then, you know, she'll be able to interject on her own. But so that's that. There is also this like bro culture, which I'm trying to navigate around. But I think that most people, most women are find a hard time navigating around it, thinking that they have to, like, become a bro to be part of this, like boys club bro cultures. So just being aware that your culture is a thing and maybe minimizing it, I'm not gonna say that you can't be who you are. I didn't want to have a bro culture that's that's cool, but make it maybe inclusive. May had a little bit of gal cultures, but there are just a few things on top of my you know some men and people in the community can do to help them with imposter syndrome.

Harpreet Sahota: [00:44:52] Last question before you jump into light me round here. What's the one thing you want people to learn from your story?

Lisa Shiller: [00:44:59] You can create the life that you want. Data science is your thing, that's great. If you want to go into an office and work, that's great. If you want to maybe work remotely from Mexico for six months, that's also cool. And that's possible. And to think about the life that you want and then work from working backwards from there. So this isn't just a Data science thing. It's an all around encompassing life thing. What is the life you want? And then work backwards from there to achieve it. And maybe it'll take a few weeks, maybe it'll take a few years. But think that's one of the most important things I could share with the community here.

Harpreet Sahota: [00:45:39] Go ahead, let's Jump into Lightning Round then. What is your Data science superpower?

Lisa Shiller: [00:45:47] Explaining concepts to people that aren't technical.

Harpreet Sahota: [00:45:51] I'm assuming you're a bit of a foodie. I might be wrong.

Lisa Shiller: [00:45:54] Totally true.

Harpreet Sahota: [00:45:56] So what is your most favorite dish?

Lisa Shiller: [00:45:59] Vegan Indian Tacos. I love making it because I love Indian food and I love Mexican food. So I usually just take whatever veggies I have lying around. I don't stick to like one type of veggie, and I just exterminate my vegetables. And then I have a curious spice that I use, powder and usually I'll throw in a little bit of coconut nuts just because it's creamy curries. And instead of butter, I use coconut oil, garlic or some onions.

Harpreet Sahota: [00:46:32] So what's the most difficult recipe you've cooked up?

Lisa Shiller: [00:46:36] Ok. I love to challenge myself with gluten free and vegan baking. I don't know why I do it myself, but every time it's so hard. Anything in that category.

Harpreet Sahota: [00:46:50] Yeah. That sounds like because I think with baking like the two essential ingredients are eggs and flour.

Lisa Shiller: [00:47:00] Eggs, flour, Butter. Yeah.

Harpreet Sahota: [00:47:03] So what's your favorite place to go on an adventure?

Lisa Shiller: [00:47:08] India. I've been there twice. I've spent several months there and it's like going to a different planet. I just feel like I have so much personal growth every time I go. And I love the culture. I love the people. I love the freedom.

Harpreet Sahota: [00:47:23] Any particular part of India?

Lisa Shiller: [00:47:25] My favorite place on the planet, I think, it is McCluskieganj. I also really love Auroville, which is in Tamil Nadu. And that's is it.

Harpreet Sahota: [00:47:35] So what's your signature dance moves?

Lisa Shiller: [00:47:38] I do a lot of ecstatic dance, which means there's no dance move. It's just like moving. However, your body is like moving. So even though I want dance moves, I think that that's my signature. It's just listening to how my body wants to move and doing like a weird look right now I'm kind of doing that where I'm like lifting one leg up in the air and then touching the ground with my left hand. So may be that's my signature dance right now.

Harpreet Sahota: [00:48:01] So what's some academic topic outside of Data science that you think every data scientist should spend some time researching on. 

Lisa Shiller: [00:48:08] Sustainability. Like anything to your sustainability. Maybe focus on one area like plastics or rubber and just educate yourself on that. And also way different. You know different companies are doing to tackle those issues.

Harpreet Sahota: [00:48:24] What's the number one book Fiction or non-fiction or if you want to do one of each. Both, that's completely OK that you'd recommend our audience read and your most impactful takeaway from it.

Lisa Shiller: [00:48:35] I go to nonfiction, which I guess could also be counted as fiction, but it's a book called How to Change Your Mind. And it was so impactful for me because it basically teaches you how to be happy at every moment, not happy at every moment of your life, but have your baseline be happy. And how to change your minds in order to achieve that. And I think that's what everybody wants. So anything of the relevance.

Harpreet Sahota: [00:49:03] Everyone like that a lot. So does that delve into like any neuroscience or anything like that about how to be happier or is it just kind of changing your internal dialogue? And how does that work?

Lisa Shiller: [00:49:16] It is a little bit of, they explain a little bit of the neuroscience behind it, a very basic level of like, you know, you suck as mindfully into one thing. It's a type of generally what pops in your brain and that's you want that. It doesn't go super deep into the narrow sense, but it's such a nice.

Harpreet Sahota: [00:49:35] I'll definitely, I'll include that book in the show notes and check it out myself. So if you could somehow get a magic telephone that allowed you to contact 20 year old Lisa, what would you tell her first? First, tell us where 20 year old Lisa was, what was she up to? And then what you would tell her at that point?

Lisa Shiller: [00:49:52] Twenty year old Lisa was in university. I think I was in second year at that point. I kind of finished my party phase and I was just really intensely focusing on math and because running the marathon stats projects at point. So if I was to call that Lisa, I was also a super protectionist. If I get any less than 100% on a test, I would have a nervous breakdown. So I think I would call the 20 year old Lisa and I would say it's OK to be wrong sometimes and there's nothing wrong with yo, everything's OK.

Harpreet Sahota: [00:50:29] What's the best advice you've ever received?

Lisa Shiller: [00:50:32] It's what I've just said, there is nothing wrong with you and everything's gonna be okay in no matter way. What ever happens in your life, in my life, anyone's life, it's okay no matter what, even if I die, it's going to be okay no matter what. And, there is nothing wrong with you and there is nothing wrong with anyone else, we are all OK.

Harpreet Sahota: [00:50:49] So what motivates you?

Lisa Shiller: [00:50:51] Going out on nature walks and just generally being nature. It is like this total reset, getting out into nature, I come back and I feel fresh, actual motivated with life and projects. Usually I'm actually problem solving in the back of my mind when I'm on these nature walks as well.

Harpreet Sahota: [00:51:11] Yes, it's definitely very common theme among a lot of great thinkers that take some time away from your office, go on a walk and just think freely. And sometimes the solutions to your problems kind of arise during that downtime, I guess, for lack of a better word.

Lisa Shiller: [00:51:28] Exactly. And sometimes in your sleep as well.

Harpreet Sahota: [00:51:31] Yeah. Yeah. What song is giving you life right now? What song do you have on repeat?

Lisa Shiller: [00:51:35] Most recently I've been listening to a lot of Mac Miller and there's a song called Duno, which is, I don't know, it's like really kind of like a cheesy and sad but super show. And I listen to it so many times, and I know other words so I can listen to it while working and not be distracted.

Harpreet Sahota: [00:51:55] Lisa, how could people connecting to you? Where could they find you?

Lisa Shiller: [00:51:59] So on my website, lisashiller.com. And then from there you have all my contact information. You can read a little bit about me. Shoot me an email. I would love to hear from you. Yeah. Let's go next.

Harpreet Sahota: [00:52:20] Lisa, thank you so much for taking time out of your schedule today. I really really appreciate you coming on the show.

Lisa Shiller: [00:52:26] My absolutely pleasure. Thank you.