 Yeah, good afternoon to one another present here. My topic is novice to ML researcher. So, who am I? I'm the research intern at Sama. So, this is my background and this is how I started my career with banking and then I was with Deloitte doing financial auditing. Then I was the founder of a STEM Atex startup for almost seven years and then I became a research intern. So, you're gonna hear about my journey. So, yeah, what was the turning point? As I mentioned in the previous slide, I'm the mentor for technology. I'm an advocate of girls in technology and women in technology. Why? We know. Yeah, you can just take a look around and see how many women are among us. It's not that women are not smart or intelligent or not in STEM, but this somewhere, you know, we need to do a lot of work to bring in to STEM. So, I was mentoring these two girls, Tanya and Mridhula, and we were one of the finalists out of 10,000 participants in 125 countries and I had the opportunity to go to Google. Yeah, I was there in the Google campus for five days and then it struck me really hard as a promoter of women in technology, girls in technology and I'm not in technology. That's kind of, you know, very absurd. So, when I came back, all I could see was all the AI in headlines. Yeah, so I said, like, I'm gonna become a technologist myself. So, it's really hard to do something and keep doing at it and it's also hard to stop doing something, right? So, yeah, this is how it's gonna look like. Having known my background, it's not so easy, but I was prepared to do it. Yeah, how many of you saw all the ML talks and AI talks in the last two days, like yesterday and today and thought, yeah, maybe, maybe I should get into ML. Can you show up your hands? How many of you think, like, yeah, should I get into ML? Yeah, I'm gonna say anyone can do ML and DL. Yeah, you can just search ML, AI in Google and these are some of the things that's gonna come up. But, yeah, first I was to my rescue and last year was my first spike and I never knew something like that existed before. So, last year was my first spike and this year I'm here talking to you guys and girls. So, fast AI, it's, I don't know how many of you've heard about a fast AI. Show off your hands. That's good and I keep telling Jeremy as my mentor. So, why fast AI? Yeah, it's a top-down approach versus bottom-up approach. So, the basic thing I did was, you know, I did an ML course by Andrew NG and then I went on to start deep learning because I saw all scikit-learn libraries, you know, do a fit, then predict, not challenging and not interesting enough and as a woman at 40 trying to get into tech, you know, right, how tough it is and with a non-traditional background doing every bit of, you know, finance, entrepreneurship. So, I thought I had to do something niche. So, I started doing, so, yeah, swimming, you learn swimming, get into the water and swim, you want to play basketball, go to the ground. So, this is what fast AI preaches and if you look at all the blog posts, right, so they teach about how to classify dogs versus pets and identify the movie review sentiments, right. So, this is fast AI. It has like seven courses for the first part one and then seven courses for part two. So, you get to learn to build a movie review sentiment, a recommendation classifier, I mean, recommendation engine building and then classification and all that stuff. Okay. So, all these concepts are demystified and then you have discussion forums, you have a share your work here. So, what I wanted to, I want to just give like few lines about what I built. I didn't want to build a cat versus dogs. Yeah, this is Sambas in Bangalore, Karnataka. Yeah, this is another bus. So, my parents had the trouble of finding the bus names when they travel outside of Tamil Nadu. So, I wanted to build a bus name boat classifier and as with all ML models, I had tough, tough time, you know, finding data. So, I started taking a small problem in my particular area in KK Nagar. So, I started hunting down for bus boats, right. So, I'll just go and start clicking pictures. I created my dataset and then, you know, I had, so this work is in my GitHub. You can just check it out and this is how I went about pre-processing the data and then I built a bus name boat classifier in under like three months and the next thing I wanted to do, I wanted to have a text classifier. So, I wanted to build a success remarks classifier. So, again, I went ahead and hunted data online. I have built a dataset anybody wants to work on, you can reach out to me. It's not made public because, again, what, there's a huge bias, right? What appears normal to me may not be appear normal to others. So, I've not made it public. And currently, I work as an ML researcher with Sama and I thank Malai Khanan and Sama team for believing in me and yeah, if I can do it at 40, everybody can. Thank you.