WiDS Conference 2024 | Mariana Moyano-Mendoza | Universidad del Pacífico

Clips
More from WiDS Conference 2024

Mariana Moyano-Mendoza

Product Management Intern - Digital Product & Design

Scotiabank

play_circle_outline Introduction to Women in Data Science Worldwide event at Stanford University

play_circle_outline From Volunteer to Ambassador: Mariana's Inspiring Journey with WiDS in Lima, Peru

play_circle_outline The importance of meeting people further along in their data science careers

Info

Mariana Moyano-Mendoza | Universidad del Pacífico

Mariana Moyano-Mendoza

Product Management Intern - Digital Product & Design Scotiabank

search

(bright music)

>> Good morning, creative community, and welcome back to Stanford University in beautiful Palo Alto, California. My name's Savannah Peterson. Delighted to be here celebrating International Women's Day with you at the Women in Data Science Worldwide annual event here on campus. 400 fabulous attendees here with us in person today, with thousands around the world in 57 different countries virtually. But most importantly, I'm excited to introduce our third guest joining us here today from Lima, Peru. Mariana, thank you so much for being here.

>> Thank you so much for having me.

>> Quite a journey for you coming up from South America,

our first South American- >> Yeah. (laughs)

>> Guest on the show. You've been a part of the WiDS community now for three years. You started as a volunteer, as an usher,

you were just telling me. >> Yeah.

>> And now this year, you get to come back as an ambassador. What does that mean? What does the community mean to you?

>> Like, if I can start by telling you like the story of how I ended up with WiDS. So I was, like, in the very start of my data science career at university, and I heard about like the WiDS community because my professors are also ambassadors. So they were doing already, like, some events with WiDS every single year.

>> Savannah: Oh, cool.

>> And I wanted to, like, know more about it, and I happened to have family that live close by, so I was coincidentally close to the dates that they were having one of their conferences, and I wanted to be here, but, like, all the tickets were sold out. And I asked if there was some chance that I could come.

And- >> Yes.

>> Yeah, and they offered me this volunteering opportunity, so I would be like an usher and also attend to the conference, which was amazing, because I got to meet so many people that were already, like, ahead in their careers in data science, and it was truly an inflection point where I was like, oh, I really love this path and I love what I'm doing.

>> How important is it to meet people a few steps further down their journey in data science, but really for any woman in technology?

>> I feel like it just gives you, like, options of what you could do in the future. Like, it shows you that maybe it's not that hard to go for that position, one job, or like, aspiring at some academic position. Like, there are so many paths, and, like, it gives you a range of things that you could do at some point. And it's more about what you want to try, what you're more interested in. So I think it really, it's inspiring because you see other people doing it.

>> Yeah, and you see them applying data science in a lot of different verticals, in a lot of different ways. Every time I come to one of these shows, I bet, I'm like you, I see a new application or a different way that AI is going to change our future, and it's really inspiring, and it's super exciting. You seem so confident on your journey now. When you were younger, how did you identify science as a path you wanted to take?

>> It was actually a pretty tough thing for me because, like at school, like we have this regular test where you are, like, before going to uni, decide which major you're going to be in. And I was very confused and I was like, "Oh, maybe I should do something like with letters." Because I'm really, like, always the social impact, it's always called my attention. So I wanted to do something related to that. And I didn't really know where, like which path, because like there are many that kind of take you there, but I wanted something at the same time that could be very practical, you know, like, for example, coding. It's really direct and you can see the results as you learn how to code, as you learn how to apply certain techniques. So I kind of find it through, like while I was in uni, I took some elective courses that allowed me to see the different courses that they were having in, like, the whole major. And also something that I really like about my university is that they have, as we are not that big of a class, you always had the chance to talk with graduates that already went through the whole program. And you always had the chance to contact them or talk with your professors and kind of see what, you know,

which path you could be in. >> Yeah.

>> And even though there's, like, a regular curriculum for everybody, you could kind of go through the directions that you like the most.

>> Yeah, and do those life design interviews with folks to see how that had panned out into their career. You mentioned that the classroom was very intimate, that helped your learning journey. You also mentioned that your major is relatively new.

>> Mariana: Yeah.

>> Go on.

>> So this is a major that's probably like-

>> What is your major, just to be clear?

>> Oh, so in Spanish, it's like the translation would something like information engineering, but we see a lot of, like, machine learning, deep learning, you know, data science topics. And it's a very small, like, I mean, they're constantly having new majors as well. But like this one is one of the most recent compared to like the other older majors like economics and administration. And my university, at its core, it's like a business school. And, you know, data is so important for any kind of business, and even for, like, countries, to analyze them and get insights that could be helpful for making decisions.

>> That's exciting. Do you feel like there's a lot of support

from your university? >> Yeah.

>> And because it is the new one, there's got to be a lot of excitement.

>> Yeah.

>> I mean, you're studying something that's right at the peak of the hype curve. It's pretty awesome. In addition to machine learning and data science, you also, well, this involves that, but you have a really interesting personal passion project that I want to bring up that matters. Here on "The Cube," we had the pleasure of interviewing the deaf and hard-of-hearing group at CNCF, the Cloud Native Foundation at CubeCon, back in Chicago earlier this year, and will hopefully again in Paris in the next couple of weeks. Peruvian sign language is a project that you are building something around. Tell us about that.

>> Yeah, so as I was mentioning before, I was always brought up to, like, social, you know, problems that could be happening at a community. And when we were having our decision for our final project, this is a Paris project, so I also have a teammate called Luis Lazarte. And we are both doing this for our final project of our major.

>> Savannah: Wow.

>> And, you know, we found in common that we really wanted to do something with social impact, and we were having like a brainstorming of different ideas that we could do, of like problems that are happening right now in Peru. And something that's going on right now is that there's not that much accessibility

for, like, deaf people. >> Yeah.

>> And, you know, there are more than 3,000 sign languages around the world. So like each country has its own, and we decided to, like, go with it. And something that we also were really interested in is not just bringing a solution and, you know, just because we like to code about it.

>> Right. >> But to understand and to have that empathy for the community, for the deaf community, and understanding if they actually need that. And we kind of, what we have framing is the system as a learning support for people that want to learn sign language. Because, you know, when you're learning sign language, it's not like you can write notes about it and like, okay, you're going to do the sign this way. Like, you have to practice it.

>> Right.

>> And because it's so basic gestural that you really have to be, like, actually doing it and, you know, having that feedback if it's correct or not.

>> Savannah: Right.

>> So we limited it to the Peruvian sign language alphabet because it's a like an undergraduate project and it's also because we only have, like, a year for making the whole thing, and we are in the last mile almost. We have the deadline for July, so we are working on that right now.

>> That's so exciting. What's the reception been like in the Peruvian deaf and hard-of-hearing community when you talk to 'em about this project?

>> So we actually had an interview with Elilire Rivaneta. We have something in, I don't know if it applies also for the US, but we have something some term called like linguistic system. And basically, it's a deaf person that also is in charge of teaching sign language. So we had an interview with her and her interpreter because we, at this point, we don't know how to communicate fluently in sign language. So we got to understand, like, the pains that they have and how it is very hard to like go to a hospital

or a bank and- >> Oh, my gosh, yeah.

>> Having to communicate, because people don't know Peruvian sign language, and there's not always somebody that could be there for you.

>> Savannah: Yeah.

>> So, and even for schools, like even for, like, middle schools and high schools, there are not enough professors that know sign language. So that's why we want something that could help, but like from our side, because it's like our commitment as community to also help others to feel more involved.

>> Absolutely. Do you think that AI, ML, data science are going to be tools that we can use for great improvements in accessibility over the next few years?

>> Definitely. I feel like especially because, for example, having a person like, very essential, it's very hard like in, you know, for the number of people that need this kind of services, like to be like some kind of translator or like to have this accessibility, it's going to be hard to have a, like, one-on-one.

>> Yeah.

>> So in that case, I feel like machine learning, artificial intelligence really helps to make it a little bit more massive and to reach out the communities that actually need it.

>> Absolutely. Have you learned some Peruvian sign language as a result of this project?

>> Yeah, I remember how to say, I always remember a lot.

>> Show us.

>> So this is October. And it's really typical because it's like you're, you know, taking something, because in October, we have this party to, like, Senor de los Milagros. It's, like, religiously. And it's kind of, like, mixed it up so you can do like, if you're, I don't know how to say it, but-

>> Savannah: In Espanol?

>> Can I say in Spanish? Okay. (speaking foreign language)

>> Oh, yeah, so, you got some, it's a beautiful parade, it's a procession behind you.

>> So it's very related to something from our country, so that's what makes it unique.

>> And I love that, too.

>> Yeah. >> Everyone would know, yeah. Oh, that's so cute. I'm going to remember that. Next time I'm in Peru and I need to do some sign language. We learned, Elle and I learned at our session in Chicago that this is cloud. We were at a Cloud conference, obviously, so I had them teach us a little cloud. Hopefully the audience can see that as well. Wow, really fun and beautiful. What a beautiful example. Thank you for sharing that. You brought up something really fun. You're very passionate about mentoring the next generation the same way that you came here-

>> Yeah. >> And got inspired. I love that you're empowering high schoolers, you're working with undergrads, you're doing datathons. You said something when we were getting prepped that caught my attention. You used an interesting data set in this last datathon, and it was Harry Potter data. What was that like? What did that look like?

>> So when we make the first Python workshop, we kind of like wanted to introduce people to the WiDS dataset, the healthcare dataset. But we saw, like, the reception of some students was like, okay, this is a huge data set. It has so many columns. So it's kind of hard if you are just starting to learn, like, to just see this massive data set and don't really know where to start. So what we were thinking is, okay, maybe we can make it a little bit simpler and, you know, something that people can relate to, it's like Harry Potter. I found that data set on Kaggle, and-

>> Amazing. >> Yeah. So I just adapted some things like, you know, it was a really good data set, like it had, it was reviews from Harry Potter movies. So you could have the, like, prediction if whether somebody would like it or not.

>> Savannah: Right.

>> And I kind of adapted it so it has missing values and, you know, kind of make it similar to the case of the WiDS dataset.

>> Savannah: Yeah.

>> So people could understand what kind of challenges they could find in another dataset and kind make the comparison of what they could do. So with that, I was very happy because when I presented it to the high school students, they were like, "Oh, okay, now it makes sense." Like, now I understand why, for example, there are two data sets column, the train and the test. And they were asking me like, "Oh, but I don't know why, like in the test, there is a missing column." And you know, part of like the challenge is to predict that missing column.

>> Savannah: Right.

>> But like sometimes, if it's something like concepts and different, you know, like I felt like in that moment, it was like they were having so many new concepts thrown at them that they were very confused. So with this Harry Potter example, we kind of simplifies things and kind of make a roadmap for them.

>> It makes it easier to talk about the pieces of the puzzle, so to speak, in a way that's when the lexicon isn't all new.

>> Mariana: Yeah.

>> Oh, I think that's really great, and I think that's exciting. I can imagine there were some silly conversations, too, if you're just listening around.

>> Yeah. >> Oh, that's so fun. And so did you feel like, could you tell that that made it easier to approach some of these new concepts?

>> Yeah, definitely.

It was like- >> I love that.

>> Yeah.

>> That is absolutely fantastic. Woo, okay. So what's next for you?

>> So actually this is my last semester at undergraduate, and we have to finish our project. So far we've done the data set of actually for the data set as there isn't any like recording of Peruvian sign language,

at that point we are. >> Yeah.

>> Because if you compare, for example, with the American sign language, there are many data sets that, you know, you have that accessibility to try on them and, like, try your model. But in the case of Peruvian sign language, there isn't anything. So we had to make it from scratch and we asked our friends to make the sign language for us. So in this case, it is a alphabet, it's not very much complicated. Of course, if we wanted to make, like, more complex sentences, we will need actual people from the deaf community to help us.

>> Savannah: Right.

>> But at this first phase that we are just focusing on the alphabet, we ask help for our friends. We memorize the whole alphabet and, you know, make them make different repetitions of the signs. So that would feed our model.

>> Oh, my gosh, I absolutely love that.

>> Yeah.

>> So you and your community have learned Peruvian sign language to train your models-

>> Yeah.

>> So that you could, wow, that is awesome. And also probably really eye-opening for your friends in terms of how to be more inclusive.

>> Mariana: Yeah.

>> Savannah: So what are you going to do after you graduate?

>> I'm planning on continuing on my machine learning path because I'm really into that topic.

>> Savannah: Yeah.

>> And I also like computer vision, so maybe doing something, but always, as I was mentioning you, the social impact. I'm very, like, brought to, whatever I do, I wanted to make it for good, like, you know, having tech for good. And I'm really, like, that's my motto right now.

>> I love that. Okay, I have three questions for you to close this out. Hopefully pretty short and sweet because this has been absolutely fantastic. You are an inspiring woman.

I'm sure there are- >> Thank you.

>> And you interact with a lot of, I mean that, and you interact with a lot of young people and people at different ages. And I actually don't like when we're ageist or we talk about someone early in their career. Someone could be older in their career and decide that they want to get involved in data science and machine learning.

>> Yeah.

>> Never a better time than right now to be in the industry. What would be your advice to someone curious about getting started or just getting started in our space?

>> You know, something that really helps is, like, finding that topic that you are passionate about. Like, it could be Harry Potter, it could be sign language, it could be anything you are really interested in.

>> Savannah: Yeah.

>> And once you have that, it's easier to, like, kind of seeing, okay, what has been done in this field? What techniques have been applied? And kind of going, like, step by step into understanding why is it being applied to it? Because, you know, in the range of different techniques that are in data science, like, ones are better for certain topics.

>> Savannah: Yeah.

>> So if you kind of limit and, you know, you have your own like thing that passionates more than most, it will help you to not feel like, you know, like overflow with information, so you'll have, like, something that you like and something that would let you learn easier.

>> I love that advice, and I am definitely going to clip that out as a soundbite for folks listening, because that's just it. If you're not into what the thing is, it's going to feel even more overwhelming than it is already. If you're applying this to a concept that you already really love or one of the things that actually the production team and I were doing earlier today was asking various LLMs to explain extremely deep and complex technological concepts to us like we're five year olds. And it was really cute and kind of hilarious, but I think it's just that. Find the thing you don't mind learning a lot about so when you're faced with that information repeatedly, you're not sick of it, and, you know, it's like how sports can teach people about statistics. Harry Potter can teach you about data science. I think that's an absolutely outstanding point. Since it is International Women's Day and we are all obviously celebrating ourselves, I do know there are plenty of people in the audience who are looking to be better allies to women like us. What would your advice be to the allies out there on how they can empower women in data science?

>> I think it would be mainly, like, listen. Like sometimes we could be even in the same room, but we are not really paying attention. And I think it's to be present and actually pay attention to what's going on in your surroundings, because the answer is there, it's just sometimes we probably are not very, like, there in the moment, and you will get the answers as well.

>> I like that. There's a lot that can be learned from listening.

>> Yeah. >> And actually processing, no pun intended, maybe pun intended. Actually processing that data, and hearing a woman's journey, I think that's fantastic. Mariana, I could talk to you all afternoon, I'm inspired. How can people follow along with the work that you're already doing right now with Peruvian sign language and everything else at school in Lima as well as with your future journey?

>> So in my LinkedIn we will be putting all the updates. Right now we are still at the dataset pre-processing, so it will take a while, but hopefully for July we'll have some more news.

>> Fantastic. Well, I personally will be staying tuned and I'm sure a lot of our guests are.

Mariana- >> Thank you.

>> Thank you so much for being here. Thank you for bringing the heat. We're grateful you're a part of the WiDS community, that you've spent all the time here, and as an ambassador this time around, very exciting. Thank all of you for tuning in to this fabulous great start to our all day coverage here at Palo Alto, California, at Stanford University for Women in Data Science Worldwide annual event here on International Women's Day. My name's Savannah Peterson. You're watching "The Cube," the leading source for empowering baddies in tech. (bright music)

Mariana Moyano-Mendoza | Universidad del Pac��fico

search

(bright music)

>> Good morning, creative community, and welcome back to Stanford University in beautiful Palo Alto, California. My name's Savannah Peterson. Delighted to be here celebrating International Women's Day with you at the Women in Data Science Worldwide annual event here on campus. 400 fabulous attendees here with us in person today, with thousands around the world in 57 different countries virtually. But most importantly, I'm excited to introduce our third guest joining us here today from Lima, Peru. Mariana, thank you so much for being here.

>> Thank you so much for having me.

>> Quite a journey for you coming up from South America,

our first South American- >> Yeah. (laughs)

>> Guest on the show. You've been a part of the WiDS community now for three years. You started as a volunteer, as an usher,

you were just telling me. >> Yeah.

>> And now this year, you get to come back as an ambassador. What does that mean? What does the community mean to you?

>> Like, if I can start by telling you like the story of how I ended up with WiDS. So I was, like, in the very start of my data science career at university, and I heard about like the WiDS community because my professors are also ambassadors. So they were doing already, like, some events with WiDS every single year.

>> Savannah: Oh, cool.

>> And I wanted to, like, know more about it, and I happened to have family that live close by, so I was coincidentally close to the dates that they were having one of their conferences, and I wanted to be here, but, like, all the tickets were sold out. And I asked if there was some chance that I could come.

And- >> Yes.

>> Yeah, and they offered me this volunteering opportunity, so I would be like an usher and also attend to the conference, which was amazing, because I got to meet so many people that were already, like, ahead in their careers in data science, and it was truly an inflection point where I was like, oh, I really love this path and I love what I'm doing.

>> How important is it to meet people a few steps further down their journey in data science, but really for any woman in technology?

>> I feel like it just gives you, like, options of what you could do in the future. Like, it shows you that maybe it's not that hard to go for that position, one job, or like, aspiring at some academic position. Like, there are so many paths, and, like, it gives you a range of things that you could do at some point. And it's more about what you want to try, what you're more interested in. So I think it really, it's inspiring because you see other people doing it.

>> Yeah, and you see them applying data science in a lot of different verticals, in a lot of different ways. Every time I come to one of these shows, I bet, I'm like you, I see a new application or a different way that AI is going to change our future, and it's really inspiring, and it's super exciting. You seem so confident on your journey now. When you were younger, how did you identify science as a path you wanted to take?

>> It was actually a pretty tough thing for me because, like at school, like we have this regular test where you are, like, before going to uni, decide which major you're going to be in. And I was very confused and I was like, "Oh, maybe I should do something like with letters." Because I'm really, like, always the social impact, it's always called my attention. So I wanted to do something related to that. And I didn't really know where, like which path, because like there are many that kind of take you there, but I wanted something at the same time that could be very practical, you know, like, for example, coding. It's really direct and you can see the results as you learn how to code, as you learn how to apply certain techniques. So I kind of find it through, like while I was in uni, I took some elective courses that allowed me to see the different courses that they were having in, like, the whole major. And also something that I really like about my university is that they have, as we are not that big of a class, you always had the chance to talk with graduates that already went through the whole program. And you always had the chance to contact them or talk with your professors and kind of see what, you know,

which path you could be in. >> Yeah.

>> And even though there's, like, a regular curriculum for everybody, you could kind of go through the directions that you like the most.

>> Yeah, and do those life design interviews with folks to see how that had panned out into their career. You mentioned that the classroom was very intimate, that helped your learning journey. You also mentioned that your major is relatively new.

>> Mariana: Yeah.

>> Go on.

>> So this is a major that's probably like-

>> What is your major, just to be clear?

>> Oh, so in Spanish, it's like the translation would something like information engineering, but we see a lot of, like, machine learning, deep learning, you know, data science topics. And it's a very small, like, I mean, they're constantly having new majors as well. But like this one is one of the most recent compared to like the other older majors like economics and administration. And my university, at its core, it's like a business school. And, you know, data is so important for any kind of business, and even for, like, countries, to analyze them and get insights that could be helpful for making decisions.

>> That's exciting. Do you feel like there's a lot of support

from your university? >> Yeah.

>> And because it is the new one, there's got to be a lot of excitement.

>> Yeah.

>> I mean, you're studying something that's right at the peak of the hype curve. It's pretty awesome. In addition to machine learning and data science, you also, well, this involves that, but you have a really interesting personal passion project that I want to bring up that matters. Here on "The Cube," we had the pleasure of interviewing the deaf and hard-of-hearing group at CNCF, the Cloud Native Foundation at CubeCon, back in Chicago earlier this year, and will hopefully again in Paris in the next couple of weeks. Peruvian sign language is a project that you are building something around. Tell us about that.

>> Yeah, so as I was mentioning before, I was always brought up to, like, social, you know, problems that could be happening at a community. And when we were having our decision for our final project, this is a Paris project, so I also have a teammate called Luis Lazarte. And we are both doing this for our final project of our major.

>> Savannah: Wow.

>> And, you know, we found in common that we really wanted to do something with social impact, and we were having like a brainstorming of different ideas that we could do, of like problems that are happening right now in Peru. And something that's going on right now is that there's not that much accessibility

for, like, deaf people. >> Yeah.

>> And, you know, there are more than 3,000 sign languages around the world. So like each country has its own, and we decided to, like, go with it. And something that we also were really interested in is not just bringing a solution and, you know, just because we like to code about it.

>> Right. >> But to understand and to have that empathy for the community, for the deaf community, and understanding if they actually need that. And we kind of, what we have framing is the system as a learning support for people that want to learn sign language. Because, you know, when you're learning sign language, it's not like you can write notes about it and like, okay, you're going to do the sign this way. Like, you have to practice it.

>> Right.

>> And because it's so basic gestural that you really have to be, like, actually doing it and, you know, having that feedback if it's correct or not.

>> Savannah: Right.

>> So we limited it to the Peruvian sign language alphabet because it's a like an undergraduate project and it's also because we only have, like, a year for making the whole thing, and we are in the last mile almost. We have the deadline for July, so we are working on that right now.

>> That's so exciting. What's the reception been like in the Peruvian deaf and hard-of-hearing community when you talk to 'em about this project?

>> So we actually had an interview with Elilire Rivaneta. We have something in, I don't know if it applies also for the US, but we have something some term called like linguistic system. And basically, it's a deaf person that also is in charge of teaching sign language. So we had an interview with her and her interpreter because we, at this point, we don't know how to communicate fluently in sign language. So we got to understand, like, the pains that they have and how it is very hard to like go to a hospital

or a bank and- >> Oh, my gosh, yeah.

>> Having to communicate, because people don't know Peruvian sign language, and there's not always somebody that could be there for you.

>> Savannah: Yeah.

>> So, and even for schools, like even for, like, middle schools and high schools, there are not enough professors that know sign language. So that's why we want something that could help, but like from our side, because it's like our commitment as community to also help others to feel more involved.

>> Absolutely. Do you think that AI, ML, data science are going to be tools that we can use for great improvements in accessibility over the next few years?

>> Definitely. I feel like especially because, for example, having a person like, very essential, it's very hard like in, you know, for the number of people that need this kind of services, like to be like some kind of translator or like to have this accessibility, it's going to be hard to have a, like, one-on-one.

>> Yeah.

>> So in that case, I feel like machine learning, artificial intelligence really helps to make it a little bit more massive and to reach out the communities that actually need it.

>> Absolutely. Have you learned some Peruvian sign language as a result of this project?

>> Yeah, I remember how to say, I always remember a lot.

>> Show us.

>> So this is October. And it's really typical because it's like you're, you know, taking something, because in October, we have this party to, like, Senor de los Milagros. It's, like, religiously. And it's kind of, like, mixed it up so you can do like, if you're, I don't know how to say it, but-

>> Savannah: In Espanol?

>> Can I say in Spanish? Okay. (speaking foreign language)

>> Oh, yeah, so, you got some, it's a beautiful parade, it's a procession behind you.

>> So it's very related to something from our country, so that's what makes it unique.

>> And I love that, too.

>> Yeah. >> Everyone would know, yeah. Oh, that's so cute. I'm going to remember that. Next time I'm in Peru and I need to do some sign language. We learned, Elle and I learned at our session in Chicago that this is cloud. We were at a Cloud conference, obviously, so I had them teach us a little cloud. Hopefully the audience can see that as well. Wow, really fun and beautiful. What a beautiful example. Thank you for sharing that. You brought up something really fun. You're very passionate about mentoring the next generation the same way that you came here-

>> Yeah. >> And got inspired. I love that you're empowering high schoolers, you're working with undergrads, you're doing datathons. You said something when we were getting prepped that caught my attention. You used an interesting data set in this last datathon, and it was Harry Potter data. What was that like? What did that look like?

>> So when we make the first Python workshop, we kind of like wanted to introduce people to the WiDS dataset, the healthcare dataset. But we saw, like, the reception of some students was like, okay, this is a huge data set. It has so many columns. So it's kind of hard if you are just starting to learn, like, to just see this massive data set and don't really know where to start. So what we were thinking is, okay, maybe we can make it a little bit simpler and, you know, something that people can relate to, it's like Harry Potter. I found that data set on Kaggle, and-

>> Amazing. >> Yeah. So I just adapted some things like, you know, it was a really good data set, like it had, it was reviews from Harry Potter movies. So you could have the, like, prediction if whether somebody would like it or not.

>> Savannah: Right.

>> And I kind of adapted it so it has missing values and, you know, kind of make it similar to the case of the WiDS dataset.

>> Savannah: Yeah.

>> So people could understand what kind of challenges they could find in another dataset and kind make the comparison of what they could do. So with that, I was very happy because when I presented it to the high school students, they were like, "Oh, okay, now it makes sense." Like, now I understand why, for example, there are two data sets column, the train and the test. And they were asking me like, "Oh, but I don't know why, like in the test, there is a missing column." And you know, part of like the challenge is to predict that missing column.

>> Savannah: Right.

>> But like sometimes, if it's something like concepts and different, you know, like I felt like in that moment, it was like they were having so many new concepts thrown at them that they were very confused. So with this Harry Potter example, we kind of simplifies things and kind of make a roadmap for them.

>> It makes it easier to talk about the pieces of the puzzle, so to speak, in a way that's when the lexicon isn't all new.

>> Mariana: Yeah.

>> Oh, I think that's really great, and I think that's exciting. I can imagine there were some silly conversations, too, if you're just listening around.

>> Yeah. >> Oh, that's so fun. And so did you feel like, could you tell that that made it easier to approach some of these new concepts?

>> Yeah, definitely.

It was like- >> I love that.

>> Yeah.

>> That is absolutely fantastic. Woo, okay. So what's next for you?

>> So actually this is my last semester at undergraduate, and we have to finish our project. So far we've done the data set of actually for the data set as there isn't any like recording of Peruvian sign language,

at that point we are. >> Yeah.

>> Because if you compare, for example, with the American sign language, there are many data sets that, you know, you have that accessibility to try on them and, like, try your model. But in the case of Peruvian sign language, there isn't anything. So we had to make it from scratch and we asked our friends to make the sign language for us. So in this case, it is a alphabet, it's not very much complicated. Of course, if we wanted to make, like, more complex sentences, we will need actual people from the deaf community to help us.

>> Savannah: Right.

>> But at this first phase that we are just focusing on the alphabet, we ask help for our friends. We memorize the whole alphabet and, you know, make them make different repetitions of the signs. So that would feed our model.

>> Oh, my gosh, I absolutely love that.

>> Yeah.

>> So you and your community have learned Peruvian sign language to train your models-

>> Yeah.

>> So that you could, wow, that is awesome. And also probably really eye-opening for your friends in terms of how to be more inclusive.

>> Mariana: Yeah.

>> Savannah: So what are you going to do after you graduate?

>> I'm planning on continuing on my machine learning path because I'm really into that topic.

>> Savannah: Yeah.

>> And I also like computer vision, so maybe doing something, but always, as I was mentioning you, the social impact. I'm very, like, brought to, whatever I do, I wanted to make it for good, like, you know, having tech for good. And I'm really, like, that's my motto right now.

>> I love that. Okay, I have three questions for you to close this out. Hopefully pretty short and sweet because this has been absolutely fantastic. You are an inspiring woman.

I'm sure there are- >> Thank you.

>> And you interact with a lot of, I mean that, and you interact with a lot of young people and people at different ages. And I actually don't like when we're ageist or we talk about someone early in their career. Someone could be older in their career and decide that they want to get involved in data science and machine learning.

>> Yeah.

>> Never a better time than right now to be in the industry. What would be your advice to someone curious about getting started or just getting started in our space?

>> You know, something that really helps is, like, finding that topic that you are passionate about. Like, it could be Harry Potter, it could be sign language, it could be anything you are really interested in.

>> Savannah: Yeah.

>> And once you have that, it's easier to, like, kind of seeing, okay, what has been done in this field? What techniques have been applied? And kind of going, like, step by step into understanding why is it being applied to it? Because, you know, in the range of different techniques that are in data science, like, ones are better for certain topics.

>> Savannah: Yeah.

>> So if you kind of limit and, you know, you have your own like thing that passionates more than most, it will help you to not feel like, you know, like overflow with information, so you'll have, like, something that you like and something that would let you learn easier.

>> I love that advice, and I am definitely going to clip that out as a soundbite for folks listening, because that's just it. If you're not into what the thing is, it's going to feel even more overwhelming than it is already. If you're applying this to a concept that you already really love or one of the things that actually the production team and I were doing earlier today was asking various LLMs to explain extremely deep and complex technological concepts to us like we're five year olds. And it was really cute and kind of hilarious, but I think it's just that. Find the thing you don't mind learning a lot about so when you're faced with that information repeatedly, you're not sick of it, and, you know, it's like how sports can teach people about statistics. Harry Potter can teach you about data science. I think that's an absolutely outstanding point. Since it is International Women's Day and we are all obviously celebrating ourselves, I do know there are plenty of people in the audience who are looking to be better allies to women like us. What would your advice be to the allies out there on how they can empower women in data science?

>> I think it would be mainly, like, listen. Like sometimes we could be even in the same room, but we are not really paying attention. And I think it's to be present and actually pay attention to what's going on in your surroundings, because the answer is there, it's just sometimes we probably are not very, like, there in the moment, and you will get the answers as well.

>> I like that. There's a lot that can be learned from listening.

>> Yeah. >> And actually processing, no pun intended, maybe pun intended. Actually processing that data, and hearing a woman's journey, I think that's fantastic. Mariana, I could talk to you all afternoon, I'm inspired. How can people follow along with the work that you're already doing right now with Peruvian sign language and everything else at school in Lima as well as with your future journey?

>> So in my LinkedIn we will be putting all the updates. Right now we are still at the dataset pre-processing, so it will take a while, but hopefully for July we'll have some more news.

>> Fantastic. Well, I personally will be staying tuned and I'm sure a lot of our guests are.

Mariana- >> Thank you.

>> Thank you so much for being here. Thank you for bringing the heat. We're grateful you're a part of the WiDS community, that you've spent all the time here, and as an ambassador this time around, very exciting. Thank all of you for tuning in to this fabulous great start to our all day coverage here at Palo Alto, California, at Stanford University for Women in Data Science Worldwide annual event here on International Women's Day. My name's Savannah Peterson. You're watching "The Cube," the leading source for empowering baddies in tech. (bright music)