The conversation this week is with Jigyasa Grover. Jigyasa recently co-authored a book titled Sculpting Data for ML: The First Act of Machine Learning. The book is a combination of a myriad of experiences from brief stints at Facebook, the National Research Council of Canada, and the Institute of Research and Development France, involving data science, mathematical modeling, and software engineering. She graduated from the University of California, San Diego with a master's degree in computer science and an artificial intelligence specialization. Jigyasa is presently applying her past experiences and knowledge towards applied machine learning in the online advertisement prediction and ranking domain. She served as the director of Women Who Code and lead of Women Techmakers for a handful of years to help bridge the gender gap and technology. In her quest to build a powerful community of girls and boys alike. And believing we rise by lifting others. She mentored aspiring developers and machine learning enthusiasts in various global programs.
If you are interested in learning about how AI is being applied across multiple industries, be sure to join us at a future AppliedAI Monthly meetup and help support us so we can make future Emerging Technologies North non-profit events!
Resources and Topics Mentioned in this Episode
Enjoy!
Your host,
Justin Grammens
The conversation this week is with Jigyasa Grover. Jigyasa recently co-authored a book titled Sculpting Data for ML: The First Act of Machine Learning. The book is a combination of a myriad of experiences from brief stints at Facebook, the National Research Council of Canada, and the Institute of Research and Development France, involving data science, mathematical modeling, and software engineering. She graduated from the University of California, San Diego with a master's degree in computer science and an artificial intelligence specialization. Jigyasa is presently applying her past experiences and knowledge towards applied machine learning in the online advertisement prediction and ranking domain. She served as the director of Women Who Code and lead of Women Techmakers for a handful of years to help bridge the gender gap and technology. In her quest to build a powerful community of girls and boys alike. And believing we rise by lifting others. She mentored aspiring developers and machine learning enthusiasts in various global programs.
If you are interested in learning about how AI is being applied across multiple industries, be sure to join us at a future AppliedAI Monthly meetup and help support us so we can make future Emerging Technologies North non-profit events!
Resources and Topics Mentioned in this Episode
Enjoy!
Your host,
Justin Grammens
Jigyasa Grover 0:00
And when the list of authors came out, they wanted us, you know, to submit little bios and pictures about ourselves. That's when I realized I was one of the youngest and probably the only one hand on the interest of the developers. So that was pretty exciting for me and I love the limitless space for imagination. I was really honored when all my contributions to open source by the Red Hat women in open source academic award I got and that's exactly where I contribute open source to introducing me to the world of machine learning.
AI Announcer 0:30
Welcome to the conversations on applied AI podcast where Justin grumman's and the team at emerging technologies North talk with experts in the fields of artificial intelligence and deep learning. In each episode, we cut through the hype and dive into how these technologies are being applied to real world problems today. We hope that you find this episode educational and applicable to your industry and connect with us to learn more about our organization at applied ai.mn. Enjoy.
Justin Grammens 1:01
Welcome everyone to the conversations on applied AI Podcast. Today we're talking with Jigyasa Grover. Jigyasa recently co authored a book titled Sculpting Data for ML: The First Act of Machine Learning. The book is a combination of a myriad of experiences from brief stints at Facebook, National Research Council of Canada and Institute of Research and Development France, involving data science, mathematical modeling and software engineering. She graduated from the University of California, San Diego with a master's degree in computer science and an artificial intelligence specialization to Jigyasa is presently applying her past experiences and knowledge towards applied machine learning in the online advertisement prediction and ranking domain. I love this part of her bio, she served as the director of Women Who Code and lead of Women Techmakers for a handful of years to help bridge the gender gap and technology. In her quest to build a powerful community of girls and boys alike. And believing in we rise by lifting others. She mentored aspiring developers and machine learning enthusiasts in various global programs. Thank you, Jigyasa for all that you do to help the next generation of machine learning engineers. And I'm really, really excited to have you on the program today.
Jigyasa Grover 2:04
Thank you so much, Justin, for having me. And I'm pretty excited as well.
Justin Grammens 2:08
I gave a brief intro with regards to some of the places that you've worked leading up to where you are today. I mean, I'm just wondering if you maybe could elaborate, give us a little bit more detail on what the trajectory of your career was, you know, from coming out of school to writing a book to sort of what you're doing today here.
Jigyasa Grover 2:22
For that, I'll have to like hit rewind a bit. And we go back to the year 2013 and 2014. I just started out my undergraduate degree. And I was just kind of like, curious as to what I want to do in my career and what I do want to do in my life in general. And that's what I came across open source and contributing to open source. So since 2014, I've been like actively working in this multinational collaborative environment, you know, adapting to variances in cultures, time zones, working styles, because being in a completely different location, and I was just like an icon on the screen. When I used to work in open source back then there was very little video conferencing going on, it was mostly like Internet Relay Chat, and so on. That is what attracted me to open source the most like the creative independence, it gave me to work on projects of my own choice to work from anywhere and not revealing my full identity. Because let's be honest, I did have kind of like an imposter syndrome there. So I was working on a project called Pharaoh. And as I said, I was just like an icon on the screen. I used to interact with mentors from China or Singapore or friends. And they were also writing like a book on the side, it was kind of like a community fat book, which was Tarot 5.0. And when the list of authors came out, they wanted us, you know, to submit little bios and pictures about ourselves. That's when I realized I was one of the youngest and probably the only women on the rest of the developers and stuff. So that was pretty exciting for me, and I loved you know, the limitless space for imagination and stuff. I was appreciated internationally by the Red Hat human in open source academic award I got and that's exactly where I contribute open source to introducing me to the world of machine learning. Because as I was working on like a project called Soucy, which was aimed at building a virtual digital assistant in open source with developers from force Asia, which is an open source community during my participation as a Google Summer of Code student. That's where I you know, came to know about natural language processing, machine learning and so on. I started taking classes in school and pursue my interest in this field I applied for you not a specialized degree, as you mentioned that I did graduate from University of California San Diego with a master's degree in a similar domain and that's where my trajectory from open source you know, building web apps Android app, turn to like a machine learning kind of specialization at present here I am applying my experiences and knowledge towards online ad prediction and ranking at Twitter.
Justin Grammens 4:53
Okay, awesome. You were talking about open source and you know, I huge proponent of open source, I mean, I've been using it for pretty much My entire career, the Apache foundation is just, you know, amazing. But even companies like Red Hat, they do some great things by giving back to the to the Linux Foundation and the Linux community. And one thing I think is really cool about software engineering is people can speak any language. But yet the code is universal, right? People are committing code, and everybody knows how to write code, but they may not even be able to be able to speak to each other.
Jigyasa Grover 5:23
Absolutely like that. Sorta like my timezone not completely off. I used to send a message. And, you know, because I was based in India back then. And then I was speaking at completely different language, one of my first mentors was from China. And the only thing kind of unnecessarily in common was like the code that we shared with each other. So definitely, it was a learning experience. And I was so glad because it gave me like a sense of achievement that I was doing something other than the coursework that I used to do during undergrad study.
Justin Grammens 5:51
Understood for sure. You mentioned the Google Summer of Code. How did you get into that? I'm curious,
Jigyasa Grover 5:56
When I was talking to my mentors from the Arab community that I talked about, they said they were working on a project, which was associated with source Asia, the open source community, and they were applying for, you know, being a mentorship program for vocational schools, why don't you apply as a student I kind of like, wrote a big application, it was mostly like a project proposal. That was also my first time writing a project proposal, I learned a lot, you know, like, what are the requirements, you have to have like a timeline ready, deliverables, and so on. So that's how I got into fourth time. And I was working on like a search, utility written in small talk, which is, again, multiple, very different languages people don't use. But for me, it was a great exposure to now object oriented languages. The next year, as I said, I started working on TLC, which is kind of like the virtual digital assistant. So the next year, I was kind of like, experienced and then applied, and I got in again, so I was like, pretty excited, you know, two times in a row. And I thought, like, why not give it a shot? Why not give it a shot, like third time, but then by then Google had changed rules that he cannot apply to the same program, you know, more than twice because they wanted to give everyone kind of like equal opportunities, which I totally understand. But I still wanted to be involved in the community. And that's when I applied to be a mentor. I mentored young students or like enthusiasts who were applying for these programs for the first time. And they're on I also ended up becoming an Organization Administrator for fire. Also, initially, as I said, Pharaoh was associated with fossasia. But then we kind of like made styro its own organization, and who quite proud because 50% of our participants were like females from developing nation. And that's what we kind of like in to have, like diverse people from different backgrounds, as I said, like the universal language was good to have in that program.
Justin Grammens 7:43
And how would you find that the females in these in these developing countries, do you have satellite groups that meet there?
Jigyasa Grover 7:49
Some of our members were kind of like professors as well. So and they used to, you know, communicate with other professors in these universities and spread the word, you know, via the mailing list in their own universities asking people to participate and submit like a project proposal.
Justin Grammens 8:04
Gotcha. Yeah, really, really cool. No, I love that idea of just giving back and getting getting more and more women into the field. One of the things that I like to ask people is, how do you define artificial intelligence or machine learning? If you're on an elevator, and somebody says, What do you do? Well, you know, how do you use this AI ml stuff to? Do you have a short description, I guess.
Jigyasa Grover 8:24
So in my mind, I always try to think of AI as kind of like exploring the depth of what we have as like human intelligence. And then we try to pass on our abilities to the machines, so that, you know, they can replay and mimic human intelligence, so that we empower machine to solve like complex problems with minimal or no human intervention at all. I see. So it's kind of like a 13. You explore your own intelligence, and you also pass it on.
Justin Grammens 8:55
I see as you're passing it on to the machine, then that can learn and then, but it's always going to need some additional intelligence being fed to it as well. All right, with that, would you agree with that?
Jigyasa Grover 9:06
Yeah, totally.
Justin Grammens 9:07
Interesting. Okay. Yeah. The other question I like to ask people is like so what is a day in the life look for somebody that's in your position, you know, as people get into this career,
Jigyasa Grover 9:17
Since 2020, of course, everything's gone remote. I'll try to work from home as diligently as possible and you know, have routine but I'll one thing that I have is like always have my team ready to give you an insight into what a to do list or for machine learning engineer would look like it includes like, you know, digging up into some data finding some trend setting up experiments, which takes up like the larger part of my day is like machine learning experiments, you know, with tweaking different features, tweaking the model architecture, and then a brainstorming with stakeholders, which include product managers, data engineers, data scientists, as I said, engineering, a new feature thinking about it, and then analyzing model performance and the major part a time model metric To product metrics, because that's a huge point, you cannot always have like, oh, the my model performs like 98% accurately or like, you know, that's the recall is not something. But that should also translate into improvement in your business metrics, especially if you're working in a company that are in a product that kind of like, brings in revenue, for example, ads, prediction and ad tracking, definitely bring in revenue for Twitter. So that is something that you have to keep in mind. And that is one of the major aspect.
Justin Grammens 10:26
Interesting, I'm assuming there's a lot of data cleaning that needs to happen. And I'm kind of, you know, setting up maybe the discussion to talk a little bit about your book sculpting data for ml, what are some of the things that you're working on in that space, and maybe a little bit of a highlight of what the book talks about?
Jigyasa Grover 10:40
Sure. So data and definitely found the biggest and the strongest foundation of a machine learning model. Many times we've seen that we even if we apply, like sophisticated algorithms, or we have like, you know, high computation powers and resources, sometimes not just having access to the right data points, you know, which could be connected to the label that we have, or to the output that we want, is the biggest hurdle that we have to overcome. And that's where, you know, having the right sort of data and you know, transforming it the right way to have the features that we need in terms of providing the essential data signal to Ahmadi comes in really handy. And this is exactly what has been, you know, one of the biggest, and the most popular campaigns going on in the AI and ML community these days is like taking the data centric approach to machine learning. For so long, we've always focused on the code centric aspect as in, like, how can we make our model more sophisticated? How can we add a new layer, you know, or just simply like, how can we add more resources? How can we personalize the efforts, but really has been the focus being on like, the kind of data we are feeding, like, how are we transforming it? How are we creating new features out of it. And that is exactly what I found during my grad school studies. And when like, when I wanted to work on any novel problem, let's say I had some idea in mind, sometimes there was no data set, or sometimes there was a dataset, which I might not be happy with. So we kind of like try to Google search, and you'll find blogs and stuff like that. But rarely did we come across kind of like a platform, which was all inclusive, like told you like, here's how you curate your own data set. Here's how you source the data. Here's how you processed it, here's how you create features, and then input into the model that most of the blogs or stuff or books that you would read, would you know start from here, say data and let's start how we can input into the model and like how we can fine tune the hyper parameters and so on. So we wanted something you know, which was like, goes way beyond like, input to the model make input to the model? We thought of it as the end product and start from the deplore.
Justin Grammens 12:46
I see. I see. So yeah, I love that the sculpting aspect of the data. What are some of the challenges and for the data that you we are working on? And on a day to day basis? Say a Twitter? I don't know specifically, are you working on ads that show up while I'm looking at Twitter? Are you looking for, you know, hey, you like this person, you might want to follow this person, you know, what, what are some of the things that you're working on?
Jigyasa Grover 13:07
So particularly for my team, I work on the promoted tweets, or the sponsored content that you might see once in a while on your feed, or like other display locations, like if you're on someone's profile, and you come across, like, embedded in their tweet, one of the sponsor tweets that shows up.
Justin Grammens 13:24
So that all that real quick decision making that needs to happen when somebody's looking at a Twitter feed. Yeah, it's gotta be very fun working on a platform of that scale, right? The millions of people that are browsing Twitter,
Jigyasa Grover 13:36
That's true. But then with these security and privacy standards coming up, you know, with Apple's App Tracking transparency and the GDPR rules, we have to be very mindful of the kind of data that we end up using. So that also becomes one of the biggest obstacles that we have.
Justin Grammens 13:52
Interesting. You mentioned about data. And, you know, no one's taking this data centric approach, which I think you're totally right, everyone's talking about the parameters to tune we need to throw more GPUs at this all that sort of stuff. This opened up in my mind the idea of sort of biased, you know, data, there's a sort of this, this whole idea around, we're learning with bias data. So now our models are becoming more and more biased. You talk about that a little bit in your book, or have you seen that in your professional career.
Jigyasa Grover 14:17
So I do have seen it in my professional career and personal projects as well. But specifically, we believe that bias and data itself is such a huge topic that we wanted to keep it more elementary and we do touch the topic, but we don't go in depth we do talk about you know, how to have like exploratory data analysis to you know, check for bytes and stuff. But biases in itself can be off you know, 100 different types are based on the kind of data that you have. So probably a you know, uh, we can work on this next or like something that definitely needs to come up.
Justin Grammens 14:53
You thinking about a follow up book that I mean, how's it been a lot of work? Maybe.
Jigyasa Grover 14:57
Let's see how that goes. For sure. I'm always looking for my next big thing to do. I'm very like task oriented in that form. So I need to have like a goal in mind. Yeah,
Justin Grammens 15:07
Yeah, for sure. Well, one of the areas that I am really getting into now is tiny machine learning tiny ml may actually go into the tiny ml summit here this next week. And the one that I think is really cool about 20 ML is, you know, you can get sensor data just really quickly from the physical world, and, and allowing people to literally hook up a sensor, maybe that's an accelerometer and be able to just sort of move their hands around and all of a sudden, boom, they've got tons and tons of data that they can start working with. So that's sort of a quick way to get a lot of data, do you I mean, in your book, and maybe through what you're working on, right now, as you guys are taking a data centric approach? Are you looking for ways to help people generate their own data? How they would generate more data to help sculpt? Or is it still a lot of open datasets that are out there that can be mined? You think today,
Jigyasa Grover 15:53
in our book, specifically, we talk about, you know, Rab as a source of data, you know, like different websites that you might have, for example, one of the data set that we talked about is how you can have the onion, which is kind of like a sarcastic website that you have, and then Huffington Post, which, you know, reports, Real News, how we can combine these two websites and data from these to create a data set, let's say, which has article headline, the article text, and if it sarcastic or not, like answer label, and then you could use that data set to not only detect, like, given a piece of, let's say, text, you can tell if it's sarcastic or not. Or perhaps you could use that data set to see or if that was written from like, which stores did it belong to looks similar to that aspect. Or you can even have like, articles from Huffington Post with their tags. And given a tech snippet, you can find out like, Oh, is it like a fashion kind of like, written in a crime way? Or like, you know, it doesn't relate to fashion and so on. So these kind of like things that we discuss in the book, like considering web as like an limitless space for beta?
Justin Grammens 17:02
Absolutely, no, that's that. That is a great, great thought. You're right, there's 1000 websites being probably new brand new websites being generated every minute, you know what I mean? And then even even inside of that, obviously, there's, I always forget the thing. But it's like, just even in the past month, there's more new data that's generated on the internet than there was in the past 10 years, right. It's just, it's accelerating, just crazy. So yeah, just scraping over all that data gives you a lot of stuff that you can use. And I'm super excited, because we're going to be presenting, I think, at one of our applied AI meetups in the coming months, so very excited to have our listeners be able to attend that and get a chance to learn more about the book. One of the other things is just trying to think about how artificial intelligence, I guess will affect the future of work. I guess, you know, there are certain API's that are being applied that some people say Oh, will no longer need a, a cancer doctor, because you know, they'll be able to read the the X rays better than anybody else. Are there any projects that you're seeing in the field that you think are interesting, that are maybe a net positive to our future to like, a net negative? You know, I don't know, being in this space, maybe you're probably reading a lot and seeing what's what's going on. I always just sort of like to ask people that question.
Jigyasa Grover 18:06
Yeah, for sure. So definitely, I've been thinking about this a lot. And I believe like, the pandemic of 2020 drastically changed the scenario of remote collaboration, we have social distancing, and of course, to majority of the world to work from home utilizing video and audio calls to collaborate. And this kind of like workforce isolation has, you know, impacted creativity at all levels, you know, starting from junior developers, who might be hesitant to reach out to mentors on text or something, you know, rather, if you will, an office that at like, you know, C suite executives. So that is exactly when I think leveraging network could be a potential solution, or to elevate these shortcomings of remote collaboration, because humans are social beings. They like being around people. And if Metaverse is kind of, like, promising to transcend those physical layers and provide like a sense of being there, without being there, I really, really think that this would change the trends of remote work and you know, the hybrid work streams that we have. And I'm quite excited for the future to learn about this, you know, quirky concept of this parallel universe, where you know, you're not no longer like, an icon or like, avatar on the screen. You're mostly like, a full blown social personality in the metaverse and, you know, working in meta workspace, despite physically not being there. So I'm quite excited about that aspect.
Justin Grammens 19:28
Yeah, that's gonna be gonna be interesting. I mean, are you are you see yourself getting into maybe AR and VR in the future?
Jigyasa Grover 19:35
Yeah, probably. But I do need to do like a lot of groundwork, like a lot of fundamentals to be learned. But if during an opportunity, first of all, I'm like trying to work on the fundamentals and then move on to those aspects. But this is definitely quite an exciting thing to even think about. Because no one would have thought, you know, like, even like, even five or 10 years ago, like who would have thought,
Justin Grammens 19:58
Yeah, for sure. And I love The aspect of the junior developers as you were saying that I was thinking about when I started my career, and it was back in the late 90s. And the internet was just really taking off and getting hot, it would be impossible for me to be sitting at home and learn just a fraction of what I learned just being in the office around people that that had 510 years ahead of me. And then I could just walk over to their desk and ask questions and bounce around stuff. So it's got to be really scary, I guess, for someone who may be coming out of school trying to get themselves into a career today. Not having that, that physical mentorship, right?
Jigyasa Grover 20:31
Yeah, totally. And also, I believe, like, there's so much that you can infer from someone's body language or their expressions that sometimes it's so difficult to write that out, like, you cannot always express the emotion, like, how are you feeling? Even in an office setting, for example, just like, last week, I went in the office, and I was like, struggling for home, you know, some things to work. And there was like a senior engineer sitting right next to me, I just like asked, and then he helped me, like, didn't even take more than two minutes, just to show me like how that works. And the exact same thing I had been, you know, trying to debug myself and thinking, you know, who should I talk to? Like, whom should I message in the group or something? So definitely, those things become a lot more easier when you're like, physically in the same space?
Justin Grammens 21:16
Yeah, for sure. I think the other thing that's interesting is, is finding good engineers that live in your geographic region is being is difficult, you know, so I own a company, we're always trying to hire people, if you want everybody in the office, which I think is a perk, but it's not, you know, 100% needed. But if it's forced upon it, you know, then you only have a small geographic region that you can find these people. And so it can be very, very limiting to a company that wants to grow and expand and bring in the best talent possible to be sort of stuck if they're requiring everybody to be in the office. So there's so there's some benefits to be able to having, like a remote workforce, I totally get it. But it's it's a fine line. Because you know, the remote workforce, it's difficult for them to be junior engineers, and yet still have them be productive or learn as quickly. So haven't solved that problem yet. I mean, are you you're probably seeing that issue at Twitter, I'm guessing or just about every company, I think is finding a tough time, finding it hard to hire somebody in the area that they're physically on.
Jigyasa Grover 22:14
That's true. And it's not even about hiring new people. Like since the pandemic started, my team has gone in different areas of the country, like people have moved closer to the families thinking like how the world almost a great spike in you know, all these bad times that we had two people wanted to be closer to family, someone like wanted to buy a house and it was more affordable in the other states or something. So people definitely in my own team have moved to different parts of the country. Even though before the pandemic, everyone used to be in the same geographic location. Definitely, it's a struggle and hoping like this meta workspace brings us all together, despite not being physically together. So let's see how that goes.
Justin Grammens 22:54
Yeah, for sure. Well, speaking of junior engineers and people that are just getting started, I mean, do you have any advice on things that as you look back on your career that and your path forward that you might suggest people books, they check out conferences, they go to things they should do as they get into this field of AI and ML
Jigyasa Grover 23:11
For AI and ML I would definitely suggest you know, having doing fundamental tried, you know, learning about not just how to apply that model, but how that model actually works, how that algorithm works. So, that is very important because according aspect it would definitely change from like, which company you work for which organization you work for which project you're working for, you know, maybe someone would you be using a different library to implement that. So implementation can definitely can because the fundamentals are important. And on the other hand, I believe like anyone who is a machine learning engineer specifically also needs to have like good software engineering skills, because this is one thing to have like you know, model in your mind or like a model coded in you know, your local notebook or something. But once you will probably grow organization or bigger project that you have to scale the service. Definitely, engineering skills come in handy and regardless of good engineering skills, software engineering skills are very important. Whether you work for like back in engineering, mobile engineering, any kind of engineering basically, also, having mentors is really important. For me, I found my mentors when I was working in open source and open source is another great place to start with machine learning if you need experience without having to, you know, intern at a place or like, kind of apply for internships, apply for full time position, apply for research, scholarships, and so on. Because there you don't need any position to work on you can just be any individual who contributes to open source project yet gain experience. With that you will be able to explore different domains to find which one you're passionate about. You can work on computer vision and natural language processing, and then figure out like, which one did you like that? And those would be a few of my tips that you know someone who's getting started in the world of machine learning.
Justin Grammens 24:56
I love the open source aspect. I don't think I've heard that enough from from most The guests that I that I talked to that are on the podcast here is, you know, I view it as a great resume builder, right? So the moment anybody, you know, when if I interview somebody for a position here at my company, the first thing I asked them to send me your GitHub repository, and then the second thing I'm looking for is, is have they contributed any projects. And so going through and contributing, it's one thing to build your own stuff, that's cool. But I do think you're right, it open source forces you to work on a team forces you to be very transparent on your work, forces you to accept feedback, you know, change stuff with pull requests, and all that stuff. So it's just like me, somebody can point to a project to me and say, Look, I worked on this. And here's the other view, you know, all the other developers that you know, they can give me thumbs up for the work I did, it's just a great way to essentially prove yourself, right? That's true.
Jigyasa Grover 25:47
And you learn so much about, you know, the working style, like when I was starting out with my first internship at the company, I realized I needed like very little assistant on like, how do you get because in my open source journey, I learned like no one commit per pull request, and every commit should be squashed into one, like every changes should be it's costly to uncommit, and so on. So for me, like that workflow, and that workstyle, you know, to maintain the repository very clean for the new contributors to come to that auto translation, any work in a professional setting?
Justin Grammens 26:17
Yeah, for sure. For sure. Those skills. Yeah, they're invaluable. You're right, no matter what company you go to, right? If you learn correct software development, processing software engineering process, it's so valuable, for sure. How do people reach out and connect with you? What's the best place for them to find you,
Jigyasa Grover 26:34
Twitter to date times is the best place to find me. So I'd be there add to get on Discord, Grover. Also on LinkedIn, you can find me or feel free to message me any questions that you have. And I'd be happy to help.
Justin Grammens 26:47
For each episode, we do hear, we have liner notes, and I'll put not only you know, link to your book and link to you, but then also just a number of these projects and stuff that you've talked about along the way. We'll find links to those on the internet. And so yeah, people can read the entire transcript and, and be able to link off to all this stuff. Well, is there any other topics or things that maybe you'd want to share about yourself that maybe I didn't touch on that you find interesting related to AI?
Jigyasa Grover 27:10
No, I guess you've touched multiple aspects.
Justin Grammens 27:13
Okay, girl. Great, well, cool Jigyasa. So like I said, I really excited to have you present to our group, I'll put a link to our meetup as well in those notes. And again, being an author, I have a huge amount of respect for that I have not yet written a book, I would love to write a book someday. It's sort of like on my bucket list of things to do. But I know the amount of work and effort that goes into it. And I think it's really, really cool what you did. I think it's really, really cool what you're doing with regards to helping young girls and people getting into this profession. So I wish you nothing but the best and love to have you back on the program. Maybe in a year or so we can sort of touch base and see how things are going on in the metaverse. Right. Maybe it'll be doing doing something cool in that space.
Jigyasa Grover 27:52
That'd be that'd be so cool. Yeah, for sure. Thank you so much for having me. And I'm excited to talk to the community in an attractive way. And I'll be talking about different approaches to build a data set, the guided and the unguided search thing we'll be teaching next week.
AI Announcer 28:08
You've listened to another episode of the conversations on applied AI podcast. We hope you are eager to learn more about applying artificial intelligence and deep learning within your organization. You can visit us at applied ai.mn To keep up to date on our events and connect with our amazing community. Please don't hesitate to reach out to Justin at applied ai.mn If you are interested in participating in a future episode. Thank you for listening