 Hey, everybody. This is Christian Buckley with another MVP Buzz Chat. I'm talking today with Rahat. Hello. Hey, how are you? I'm doing well. So, so folks that don't know you, who are you? Where are you? And what do you do? Oh, my name is Rahat Yasir. Where am I? I'm in Montreal, Canada. Montreal. And what do I do? So I'm a Microsoft MVP in AI category. It's been like eight years that I'm an MVP. First three years was in a Windows Dev platform. And later last five years in AI category. And they're like maybe two AI MPPs in Canada and around 120 something in the world. So I'm one of them. Now did you kind of switch like what you focused on or did the category kind of evolve and change underneath you? Okay. Yeah. So because my expertise is data science in AI. So at that time when I was like, when I became like the MVP in Windows Dev, I was focusing on application that are using AI and data science. So they put me into the like a Windows Dev category for that. But whenever the AI started five, six years ago, that is the time like they hooked me on that one and moved me. Yeah. Well, I know with so much of Microsoft, I've talked about this in other episodes of the other MVPs about how, you know, there was a shift with within Microsoft and the product teams a few years back where they kind of went through a wave where they laid off a lot of S debts, a lot of the testing folks and but at the same time we're hiring data analysts and to come in. So data driven product management decisions like what are we actually seeing? What are people actually using? What's the feedback on products releasing something? Look at the data. Look at the analysis there. Sometimes rolling back features or here's why we made the decision on moving this button on the screen or what what have you? Yeah, in a traditional world, what happened, what used to happen is product managers, pro APOs or executives, they used to take the decision. Yes, they were visionaries. They were setting up the vision. They had used to have some market research, but again, like it was coming from them. It's kind of like centralized decision making. But then what happened, they got to know about like, okay, we can collect all those data. We can capture everything. We can find out pattern. We can actually know how clients and customers are evolving and communicating which features they are interested about, which features they're not using at all. So then they started to use them and started to like design products based on the data, data driven product designing. Yeah. Yeah. Well, you know, one thing is that so I have two of my adult children that are both in the there's both STEM sciences kids and both of them have gotten a passion for for data analysis and both of them started playing with Power BI and other stuff. Of course, one of them nerds out. He's like, I love Python. I'm like, you love Python. Come on, you know, but it's but it's interesting how you know, data science is really touched every industry multiple roles within that. So it's a skill set. I was pushing my one of my sons like, Hey, you're going to go pursue this. This is he's actually a we call him a weather boy. He's about to graduate in atmospheric sciences is degree. Okay. And I tried to convince him like do a minor in data science like go go find something and he actually said to me. He's just like, ah, no, no, I'm okay. You know, and he's just like dad. That's one of my regrets is that I should have done the minor. I should have focused more on that. So, but he can always learn. Of course. Of course. Yeah. So the way I see it, like like last few years, still a lot of people in the within the industry were thinking like, oh, data science is a fancy thing. AI is a fancy thing. Just a buzzword or hype or something like that. But no, if you look at like, again, like we are not going to get into the argument of like statistics, machine learning, all those kind of things. It's end of the day. It's just the same finding the pattern with the. Let's get into the argument of the data science versus machine learning. I want to understand. No, I know, but the way I see it, it's kind of like in future, because of the democratization of the AI and data science, it will be so easy so that anyone will be able to use it. So how if you look at the schools, so they're like a teacher after the exam, they find out, oh, top 10 percent of the score is this top 20, like less than 20 percent is this bottom 20 percent is this. So they're using very basic stats. You can you do the same thing along with the behavior of students. They're learning pattern to find out which student needs more attention. How can they do those personalized training and learning pattern? So that is kind of like getting like a kind of like aligned with the overall learning pattern coming up with different types of like adaptive learning like strategies so that the learning is designed for each an individual. So it's not going to be just generic. Now are aren't now what you're working on aren't we just increasing the speed at which we achieve the future, the idiocracy? Not really. You've seen that classic film. It's an important film. It's always a historical document, you know, it is, but they're good and bad side in both cases. You know, of course, think of all the soda pop would get to consume in the future. I mean, awesome. Yeah, yeah, for sure. But from work wise, so if I just like to give a quick intro, it's been like quite a few years. I mean, this data science and AI domain and currently I'm working at Isaac, Isaac Instruments. It's a company based in North America. Heading offices in a Saint Bruno Montreal. So we are in like, we have offices in Toronto, Calgary, Ohio and US and we design AI solution like hardware and software solution for truck fleet. So this is very interesting the way things Isaac does. So they're one of the biggest in Canada right now. So what they do is all those like big trucks that we see they install an IoT device, a black box. They call it in metric on that one. That has tons of like sensors. Those sensors are collecting data every seconds, every milliseconds on that level. And then they have different types of like a tablet app, mobile app, web application, desktop application to help the truck fleet for their management, how the order is happening. Where are those trucks? If there is any issues or not, then we started the data science and AI. So then we saw that we're literally receiving 40,000 rows of data like 40,000 records every second from every single corners of North America and those data has like GPS information, our truck information, how the supply chain is moving. And then we started to do design like a full cloud platform in the in Azure. They were already in the cloud, but we started to like make it more robust, scalable, made it like more like better for the data science and AI related consumption. Then we are doing a big data analysis. So in a day we receive literally like billions of records like of data. Now we are kind of like finding the gist of that data summarization, aggregation, what is actually happening so that we can apply those summarization of the data in business like automation, different types of like efficiency and also find out which truck is having issue, which driver is having struggling to drive in different types of like road condition. Then the next phase we are getting into is AI. So we also hired a bunch of like data scientist and AI engineers to start to like design AI models, which a truck is going to have issues, similar kind of things, behavior, driver behavior analysis, adding different types of correlation. So that's what I do with my team, designing all those like cool data science and AI applications and they will hit the market soon like because we already did a lot of like groundwork so that our clients can enjoy the products. That's very cool. Are you working with specific industries or certain areas or is it we have customers across the board? Across the board actually because thinking like that, like if you are like a big truck fleet company or like a big Walmart, Canada, you have hundreds of trucks or any other like big truck fleet organization. You're bringing getting all those things in different ports because the shipments is coming to a big ships and then from those ports, you are literally like loading everything on that truck and delivering it into different places of within the city or in the country. One interesting thing like as I'm not from the trucking industry, the first learning from me was we're literally everyone is very connected with this trucking industry. We we don't even realize because suppose in Montreal the salad or that we eat it comes from San Francisco. So a couple of drivers take a big truck every morning like almost every day drive from Montreal to San Francisco, California, get everything, all those like a salad and everything and then drive it back and that truck has obviously fridge and everything. And for people that don't know too, that's like a four day drive. Yes. So yeah, it's yeah. Yeah. And so it's very connected and that's what we are trying to do is make that supply chain operation easier and more efficient, save money, save CO2 or CO2 emission is also like extremely important, green energy is important, saving the full is also important. So that is our responsibility with data science and AI to solve. You know, it's fascinating to me like that. So I early in my career about the first, you know, 12, 13 years I worked almost exclusively and so I was in telecom for a number of years and went to work for an early SaaS company in 2001, but I owned these massive databases. All this like I've done, you know, GIS data. So, you know, geolocation data that we've integrated into the platform and at the phone company. Here's an example of, you know, it makes a big difference when trucks show up to dig up a street to fix wires that the accuracy of the map is within like three meters versus 30 meters. That's a lot of street that you might dig up before you find what you need to fix. And so we were constantly updating that. What's interesting though with a lot of the, you know, with the IOT and you see a lot of the developments around it is that the system is actually able to go in and based on, you know, it's able to predict more accurately where there are errors within the system and things to go and look for and then where to go and you know, focus and proactively look at and say, Hey, here's the different, you know, situations, whether patterns or changes that most likely cause like a risk assessment on Hey, if we make these changes, there's this activity around these devices. They are more likely to fail. And so you can have make sure that I've got the parts of the systems and the people ready on hand in case we experienced those failures and yeah, I mean, it's there's so like, like I was saying, I mean, there's so many opportunities for AI and what you go in and automate to, to, to go and have a, you know, a better real time picture of what's happening, whether you're like, I have, I went from telecom into, you know, really kind of the, the high tech manufacturing space and sort of learned about that and we were building again, massive databases to better understand like demand planning, like we're going to build a product. If we change a design, what's impacted that so that we could guess if I make this design change, here's how it's going to affect when trucks can be on the road to stores with those modified devices. Like we'd know everything around that activity and it's automating that so that the system is doing more thinking about it. Just speeds up that process. Yeah. Yeah. It's a fascinating. It is fascinating and the way the overall global AI market is growing, it's scary actually in some cases. Because I was just for one of the report, like I was just like looking at some of the blogs and articles to put it in a presentation. How big the AI market is right now from all those like Bloomberg Forbes, like prediction. So I was looking at the number like today, the global AI market is almost 60 billion dollars, but by 2028 it's going to be 422 billion dollars. So year over year growth is 39.4 and that is massive. And one of the interesting part is 43% of this overall revenue or the growth will happen in North America, US and Canada. So there is a big chunk and it's also like a big responsibility for us to bring those innovation so that we can bring that kind of like automation as well. Yeah. There's it's a great space to get into. I mean, so you've been doing this long enough like what is your recommendation? People that want to, you know, get into this like where should they get started? What should they start experimenting with to learn more about the space? So there are like two types of two ways to get into data science and AI. First one, I see if you're already in one of the industries like any kind like teaching, like pharmaceutical biochemist or like any any other physics. So you have the opportunity or even like insurance, suppose actual science. You have already the domain knowledge. Now the next thing that you can do is learn coding, Python coding. What kind of like data science like a model that how can you design those data science model? What kind of models are there like classification, regression, time series based models? Very basic. So your domain knowledge is really useful and valuable and at this that extra skill of Python coding some like basic lifecycle of AI model development and design those kind of like data analytics products and AI products for your industry. That's one of the use case and the second in the second use case. I see if you're someone from computer science and already knows coding that is perfect. Go and learn Azure as your mission learning how to train the model. What does it mean by a model? So how do you monitor that model? How do you deploy that kind of like AI model? What is the use case so we so that you can help different kind of like domain experts and different industries and take and design the model with that much big data and then deploy those modeling for different apps of business purpose. Yeah, it's I know that there's also just tons of training. There's free training. There are certificate programs and things out there. There's a lot of ways to go in for beginners. You can have absolutely no coding background and not be in a tech, you know, field at all and still be able to go in and learn about those things. I like you. So as a guy who has two marketing degrees, of course, I've been in tech my entire career, but you know, so I understand a lot of things. But like I, you know, back in the day, I did like I knew a bunch of Unix that really lasted long, you know, but you've never been a coder. I've never been an engineer, but to get into it. But I found a couple programs like via, you know, through LinkedIn Learning and through some other resources was able to go and take some do some basic things and could have gone and pursued that. So there's a lot of opportunities to learn about that for somebody who's already in the space that that's like what what was kind of your path to becoming an MVP? I know you've been in this space for a while. You know, because I think this is a growing area even for MVPs. I think there's plenty of room for new AI. Yeah, so the my journey was a bit different because 10 to 12 years ago, I got invited to attend Microsoft Student Partner Summit, MSP Summit in Seattle. So I went there. I was loving Windows Phone and at that time I also like participated at Microsoft Imagine Cup like and then which was in Windows Phone application with your phone platform. So which kind of like eventually took me into a situation where oh, I want to do this. I want to design that, but there is not enough documentation, tutorials or something like that. I started to write tutorial by myself. I was fixing this issue and then I was writing a blog or tutorial or a video like that time which and I was publishing them on C sharp corner. It's like a technical side and later I saw that it got like maybe million plus views and everything was like I launched everything for free. Someone gave me an idea like, Hey, why don't you publish a book? Then I wrote a book on Windows Phone 8.1 application development that was the first book on that platform, which eventually transferred into Universal Windows platform application development. That was also like one of the first books that I wrote and then launched it for free. So whoever wants to download it, just download it and use it and develop a Windows Phone application. I was I was in love that much on that with that platform. I have to say, I missed the Windows Phone. I was a fan of it as well. Oh, a lot. Yeah, I miss a lot. Yeah. It's for those that never use a Windows phone. I mean, those I mean, it was a just a great UI. It really was and I'm really. Sad and I was on still on Verizon platform. So like we didn't even get the updates, you know, the AT&T people in the US, you know, got another update of newer phones like before we did. But yeah, I missed that platform. Yeah, eventually Microsoft contacted me. Hey, you published this book, that book. So we are giving you rewarding you with the MVP award and that kind of like started the journey and eventually as my domain is data science and AI more into that. I worked in research scientists and a lot of other AI places regarding AI projects. Then they rewarded me change my like overall AI MVP category into AI. So it's a very interesting journey and I still like a right blocks. I still like a gift. I give a lot of talks actually. That's what I like enjoy the most kind of like sometime I talk to business audience. Sometime I talk to like IT audience or like completely technical. Sometime I talk to a mix of both so that I can show that how to design a complete data science and AI platform where without having 10 different tools within the ecosystem or juggling around from a single place in Azure, you can do everything like story or data, build your operational databases, build your analytical data side, then do the data science training because your AI platform and the data platform needs to be closed. If it is too far, then it's it becomes very expensive because you need to move the data for different types of training like approaches. Ethics and responsible and like security. Those things will be hard to ensure as well because you need to like data movement, data balancing. Those are hard. So that's what I do sometime and sometime I even like like give presentation on new topics about how IOT we can design AI on large volume IOT like applications or even like design AI with no code, no code with the power platform Azure AI mix of both. So it's fun and every year I try to keep like 10 to 15 talks. Sometime I travel. It's also the fun part in different corners. Maybe next trip will be in Toronto in October, the November I will be in like California, San Francisco for AI dev give kind of like a maybe one of the keynote kind of like a talk. So that will be fun as well. So it's a fun ride. Yeah, there's always a lot going on. And of course anybody that you know, I'm sure you do the same thing, but it's like there's enough user groups and other just purely virtual events that are always looking for speakers. And I wish more would well, I would say reach out through like the mvp.microsoft.com site to look for speakers. I would say that, but it's not exactly easy to navigate and find people that are we almost need to have as part of that tool is kind of a speakers bureau for mvps like, Hey Microsoft, if you're watching what about a speakers bureau? That would be fantastic to have there. So people could go and find us based on the region or based on our interest in various topics and all our specialties. I think it would be fantastic, but anyway, we can we can dream, you know, Yeah, but also like it is kind of like we do it on a voluntary basis. So we have our own day job family and life and everything, but it's still like all those mvps that I see in US, Canada, Europe, everywhere. They're so passionate. They're trying to like help everyone or come up with like a new ideas, sharing new ideas. This community is amazing. I I always learn so much from everyone. Well, with all the stuff going on, maybe this is a great. I don't think I've asked anybody this, but so how do you battle with all of that and the community activities and your job, your family, the rest of things that you try to do. You know, how do you stay keep from getting burned out? For me, it's easy because things I do in the day job. I just I can just like talk about those topics at the conference as well. So I don't need to like take too much of a presentation or preparation to create the material or something like that. So it's it's pretty easy, but it's also true. Like there is like a risk of like getting burned out. That's life. Nothing to do, you know, we all are juggling with that. Yeah, it's a I mean, it's always great when you're able to your passions for the technology, the topics. I mean, similarly, I'm able to talk about things that things that I do from a community standpoint are supported by my company. I'm talking about things that are relevant to my job to my company. And so there I found that mix that wasn't always the case. I've worked other jobs where the things that I did with community were completely evenings and weekends and you know, and had to even fight for that time. But it's you know, hey, we do it again. You do what you're passionate for and sometimes you need to be realistic about what you can do and what your family can support and kind of ease off on it. But that's there's nothing wrong with that back talking with like people about try to get them to be guests on a podcast or speakers at an event and they're just like, yeah, I've just got too much and I'm just like, hey, no problem. Let's talk again in six months. You know, don't worry about it. You don't need to make excuses for that. So it's it's important that we we not piled too much on each other. If yeah. No, for sure. And even like from the like MVP and contribution level. I have noticed like our CPMs comedy program managers of MVP program. They're also very supportive. So if any year I felt oh, I didn't contributed enough. I should have like contributed more. I have communicated and like told them like, hey, I'm busy with my work or family. These new things are happening. They're like, okay, like they review your contribution of last over 4, 5, 6 years. It's as it's already 8 years for me and they're always trying to like come up with like different types of suggestion advice. Oh, put your more effort in this one so that like you feel more like less pressure and like more like involved with the community. So that always helps. Yeah, that's that's a great tip to for anybody that's a new MVPs especially is is to have a good relationship with the CPM so that you have a conversation for exactly those reasons because hey, life happens. Everybody understands that. I think the one positive that came out of the pandemic is that I think that we people in general my my observation is that people are more empathetic to others and and some of those things like you remember like if you've ever been on a call and the dog started, he's like, Oh, I'm so sorry or I'm a few minutes late and I remember being chewed out by by a VP. We get somebody who was on a webinar dialed in. I was supposed to do a demo and walk through and there's construction happening. They cut through the fiber. I lost my internet was out and so I dialed in tried to call people by the phone saying like my internet's gone. This guy like chewed me out was so angry went to my boss to complain about that. And I'm like, wow, what a jerk. Oh, it's like, it's like, it's like, it's like, it's like, it's like, it's like, it's like, I'm sorry. You know, but yeah, I think people are a lot more empathetic to those kinds of things and and work life balance as well. So yeah, that's great. Well, Rahat, so people want to find out more about you or get in touch with you. What are the best ways to reach you? LinkedIn. You can just like search my name Rahat Yasser. You can find me and follow me. These days, I have started to like share more about like as your AI different types of opportunities with the community. So every Friday, I'm trying to like post a bunch of like things kind of like learning materials so that people can go and find it and like learn something. Suppose last Friday, I I saw that Gartner published that Microsoft is the leader and also like the most stationary in this cloud AI space. So that also like I created like a nice overview use case. If you want to learn this visionary and like leader tool base. So what do you need to do? So these are like all those like links and things that you can go and like learn and like excel in your career. So that kind. So the week before I also like presented something regarding like as your MLOps because MLOps is something that everyone is also like struggling about in the market. So how to build those things and how to design it. That's that's one thing I prepared. That's very cool. You know, there I know a couple of MPPs to that have built out their blogs, their websites as resources training within specific categories found all the all the guides put all the resources just made it kind of a hub for everything around that. That's a that's a great utility. It's a good thing. If you're looking to kind of differentiate yourself and maybe pursue becoming MVP is to become that hub that go to for a category of technology. So yeah, it's like you don't need to like write the same article that is already written by someone else just promoted give them the actual credit and always give them credit. Yes. Yeah, for sure. That's more important than anything. Well, we're hot really great to meet you and hope to see you in person and event sometime soon. But you know, thanks so much for participating today. Thank you so much.