Julian: 0:00

What I find a lot with technical people like you and and even me uh and so many of the technical people I I know that just have this real passion for the technology, it started young, right? Whether it be you know like your your curiosity with the calculator and just how this device could suddenly do all that kind of stuff for you that used to be manual. I had to code stuff on a Commodore 64 console way back in the day to play a game, you know. So I just I I love a lot of people who grew up with this tech and and it's just inspired them to do more with it. Hello and welcome to the Pie Bites Podcast, where we talk about Python, career, and mindset.

Advert: 0:38

We're your host, I'm Julian Seguira, and I am Bob Belderboard. If you're looking to improve your Python, your career, and learn the mindset for success, this is the podcast for you. Let's get started.

Julian: 0:50

Hey everyone, welcome back to the Pie Bites Podcast. This is Julian, and I have a very special guest with me today. This is Asif, and this is our very first university student from North America. So I'm very excited to have you on the podcast at Sif. Asif, welcome to the Pie Bites Podcast.

Asif: 1:07

Yeah, thank you for having me here, Julian.

Julian: 1:11

It's my pleasure. So, everyone, a Sif is here from Northern Arizona University. He is a teaching assistant for Advanced Python. Uh, and we we connected late last year and we're chatting and everything. But the reason I want to bring you on here, Asif, and what and everyone listening is that he's a data scientist, he's into machine learning, he's doing stuff with AI agents, he's well, Advanced Python, as you heard, because that's he's got a job as a teaching assistant doing that at the university. Um, he's helping students all over the campus that are doing that course, and it's just incredible, really inspiring to me to see someone, and I don't mean to throw age at this, but to see someone at your age and so relatively junior compared to people that I normally host on this podcast doing such incredible stuff. And it makes me think there's no excuse for anyone to not be doing things when someone as as junior as you, and I use that term very loosely here, um, junior relatively speaking, doing so many incredible things and building such cool stuff. And so uh we're gonna dive into that today. And I'm so so excited to have you here, Asif. Um, especially after our multiple chats. Some of our chats last year and and earlier, you know, a week or so ago, I thought, damn, if we could have recorded those, that would have been awesome. But um, we'll go into this. So uh with that introduction done, Asif, do you want to introduce yourself? Who are you? What do you do? Um, tell us about this teaching assistant stuff, um, and we'll dive into it. Go for it.

Asif: 2:44

Hi, listeners. I'm Asif, and I'm from I'm from Northern Arizona University. I got graduated in the last student 25 December. And when I call it an experience, I work as an AR or ML engineer in Northern Arizona University with the well-scoped index and uh teaching assistant for advanced Python, and I'm my favorite professor who is OA. Yep. And that's all about me. And I uh starting from five years back from now, I'm always in relation learning pipelines and automated systems, and also helping students to learn Python.

Julian: 3:24

That's cool.

Asif: 3:24

That's pretty that's pretty much about me.

Julian: 3:27

That's you, that's you in a nutshell. No, I love it. It's it is very cool to see you making the effort to go beyond just the learning to help people. Um look, you know, as you know, Pybites, we're in the education space with Python. So this this is a question from me uh before we dive into your projects and everything that you're building. As a teaching assistant, you know, you have to help students, you have to sometimes take over classes or lectures and everything like that. What would you say, you know, these days, um, you know, 2025, 2026, what are the what what is the biggest challenge as a teaching assistant? What do you think is the biggest uh challenge to the students that are learning? Yeah.

Asif: 4:12

In the field of learning, like a teaching system for different kinds of students who come from different regions in my country in my college. The most difficult situation with the students to stay with you. So people got so many resources, you know, different kinds of platforms to learn Python, YouTube channels from different kinds of uh Udami courses and different kinds of Coursera courses, and different kinds of platforms such as PyBytes. People can get professionals from that, but they have to choose you because you know they have to stay with you because they learn better. So yeah, teaching from the basics and going like the most challenging situation in students get adapt to different kinds of learning techniques. Some people learn faster, and some people go slow, and some people at least don't care. So it's all about dealing with different kinds of students and making them learn and become an advanced learner. That's a big expert. That's that's a big kind of uh challenging situation to deal with the students, like help them. Uh the correct code you have to learn. This is the syntax you have to buy hard, and this is the way you have to think about a problem kind of things.

Speaker 1: 5:30

Yeah, yeah.

Asif: 5:30

When when you get good in it's when you get habituated in this, so you make students impress and come to you rather than going to forms. Yeah, but that keeps keeps you in a job. But but the most important part is although the main goal where they get the knowledge from, but they need a uh some kind of practice sessions, like exposing to the open and see how students are dealing with that. And yeah, this Pie Bytes platform makes a big difference here. Although they learn in the Pybytes, the Pybytes platform makes it a better uh ground where different students come from the different regions and experiences and how they're learning and how they're gonna implement these kind of algorithms in the real world scenarios. Yeah.

Julian: 6:20

Thanks, man. Yeah. That was uh not intended to be a plug for pie bytes, but I will take it. So thank you for that. No, that's that's cool. It's kind of nice to see that it's the challenges that you're facing as a teaching assistant, they're if they're quite timeless, you know. Uh these are the same challenges we had five, ten, fifteen years ago. How do you keep people engaged when they come from different backgrounds and how do you support them at the different levels of learning, right? Um, and it's so it's kind of nice that they have you um to be able to come to, and it kind of goes with our philosophy of one-to-one, right? You can do one-to-many, teach one-to-many, but then when people don't understand who do they talk to? So they need that one-to-one bridge. And uh it's cool that you get to you get to do that for for students in in the university. That's awesome, man. Um, so you did mention that you graduated at the end of 2025, is that right?

Asif: 7:12

Uh yeah.

Julian: 7:13

Okay, what was that? What was the degree?

Asif: 7:17

Oh, it's a master's in computer science. Completed my bachelor's in 2024 in the in the same major computer science, and I was so fascinated about this automated systems and this Python. So I then climbed up the master's ladder and finished it happily.

Julian: 7:34

Wow. With a good aggregate CGPA. Congratulations, man. That is that is no mean feat. One of my mates is going for a master's in CompSci, and I just wow, um, incredible. So hats off to you for for getting that done in a year. That's that is absolutely amazing. Um what was the what do you think? Uh you know, before uh so we we have some projects to dive into that we've that you've been building, but what was the the most challenging part about the degree?

Asif: 8:02

The most challenging part of my degree is learning new techniques, which you learn in your bachelor's, improving your skills. Like master's is nothing about improving your skills, which you learn in your bachelorate or something. So here it's a song, and you have to work on real-world data examples. So you have to in uh in bachelorate, you'll just focus on the projects which are like a kind of to the ground. But in in masters, you have to apply those techniques into the real world data, developing systems. So that that seems messy in the in in the beginnings, but later on you get habituated to that. Yeah, and you shouldn't, and yeah, you should never stop in the middle of the journey. If you're if you step into the masters and if you're afraid of Python implementing it, uh it's never it's nothing like that. You just have to try, fail, try. Masters is everything about trying and failing and again building ourselves. Yep.

Julian: 9:02

Cool. No, I love it. I I didn't do a master's, so I I I really appreciate that context. It's really nice to hear. Um, it makes it seem less daunting, so that's pretty cool. Um drinking my coffee out of my pie bites mug.

Asif: 9:20

So yeah, I I should go run for this.

Julian: 9:24

I'll have to send you one. All right, there we go. Um, okay, so look, with your master's Compsci, you wanted to specialize in data science, machine learning, AI, these kinds of things. So uh let's go into the why. So, what is it that drove you and and pulled you to this side of Python and programming? What made you go, you know what, data science is for me, because we have a lot of listeners who want to become data scientists or who already are data scientists. So I'd love to hear it from your perspective. You know, what is it that pulled you into this field? Um, because this it will directly lead us into the stuff that you've built. So tell me, tell everyone who's listening, what drove you to go down this path?

Asif: 10:02

Yeah. Yeah. Talking about this would drag me back into my which it's like I'm always fascinated about this automated systems. Fine. When my Max professor asks, what is two plus two? It takes me a minute. Like when you're when you're schooling, it's like a second class or third class or fourth class or something. It's like, yeah, you will do hands-on two plus two. Fine. Then when I when when my when after my schooling, I just used my calculator. I just pressed two plus two, it actually displayed my answer. It it doesn't need my hands. And I'm always fascinated how this automated systems, how this technology works. Fine. If I press a button, it just gives me an answer. Nowadays we'll see if if I just give a question about just something, it just directly gives me an answer. Not a not an answer, an exact detailing about the concept. Fine, I'm a robot using this, but I want to try that. I want to build one of them. Yeah. So this row into this field, and I'm always happy to share that as my memory because that's the point. Fine, as you've learned this automated system, you'll know how to do that. Fine, you'll get a you'll get a respite, and you'll get an you'll get an highness after you finish the work.

Julian: 11:25

Yeah, yeah.

Asif: 11:26

That's good. Yeah, I'm always fascinated into this automated systems. Like fine, after after this, I've come, I have come across the machine learning pipelines. Fine. You build a model, data, you take the data and you clean the data, and you apply different kinds of models for the data. Fine. I was like, not happy because you just develop a model which is of no use. Fine. This is where an automated system comes from. The machine learning. Fine. Take that model and deploy it to someone's website and give it to someone who can use that for their revolt project or a business or something. And this draw this draw my attention to build an automated system using this ML pipelines or our models.

Julian: 12:15

Yeah. That's that's cool, and that's interesting. Uh because you're talking about. So, first of all, actually, I love the the story of your childhood and how and and what I find a lot with technical people like you and and even me, uh, and so many of the technical people I know that just have this real passion for the technology. It started young, right? Whether it be, you know, like you your curiosity with the calculator and just how this device could suddenly do all that kind of stuff for you that used to be manual. Uh, but I know people who started on, you know, I had to code stuff on a Commodore 64 console way back in the day to play a game. I had to program the game into the console, you know. So I just I I love a lot of people who grew up with this tech and and it's just inspired them to do more with it, which I think is a nice driver uh to keep people motivated. But um to your point, you know, you can create it, it's that's an interesting point that you make. When you're in data science and machine learning and so on, you can create these these models, you can create these these pipelines and everything like that. But unless you apply it to something usable or appropriate or there's something useful for people, you kind of stop and you kind of think, well, what was the point? So that's really cool. So you this this whole concept of okay, I'm gonna deploy this for someone to use on their website or whatever it is. Have you done anything like that? Have you built anything around that?

Asif: 13:47

Yeah, yeah, I have built uh reliable AI agents which work for the website as a chatbot or receive their emails and summarize them and give them like in the next day, and the owner wakes up and sees, yeah, this kind of uh like your bookings. Because in order to, you know, for the businessman to see the data, it's like, yeah, the day bugged and the rooms say bugged, it's like messy. Fine. In this in this automated age, like in this automated generation, you need to wait. Uh the businessman just wants to wake up and read, yeah. Like in the natural language, he just wants to know, yeah, he's got bugged because he's not a technical person. He's not he's not a person to look up at the Microsoft Excel or the sheets or how many yeah. This really drew my attention to build an AI agent system. There are many tools in this uh system. It's like one of them is called N8N, where you can develop an when you can develop an AI agents without any code base. And you can just deploy with the server.

Julian: 14:56

Wow. Yeah, yeah. That's cool. We've we've touched N8N before, but I really want to I really want to deploy it in my garage and have a play with it. I haven't I haven't done it yet, but I will. So just to clarify, in this example of something that you built, right, you you're taking this information from people on the website using a chat bot. Um, this is for you said room booking, so you're talking about hotels.

Asif: 15:20

Yeah, it could be for any kind of hotels or different kinds of uh restaurants, how how much business they had in the last day or the last month in the natural language. Because a large language model can summary, it can take up so much of data into that. Yep. And it's how it really works. That's so cool. But they just can summarize and talk with the person like a human. So that's why it is famous first. So it could be really good for the different man and the restaurant owners who are not very comfortable with the technical things like Python or what is in it, and what is it doing to that and everything and stuff? All so he can just wake up in the morning and know in the natural language, as his friend is saying, or the person who is working at motel or the restaurant is saying, we had this much of business, and we had we are going to do this much of business today. According to according to that, they can plan their long-term and short-term goals for their kind of different businesses. Yeah. That's cool. That's true. Building different kind of AI agents. Yeah.

Julian: 16:29

Yeah, I think it's really cool. And it's it highlights a gap that I see with all this. Everyone says, you know, you can use, you know, LLMs and these code generators like lovable.dev or whatever it is, um, and all these other places, right? And and even N8N, and you can use all these technologies to just build incredible things and and save the world, right? But I think there there is not even, I think, there just is a technical barrier to all of this. And back to the original reason why you got into tech, this curiosity, you kind of have to have it to be able to push through the technical barriers. Because, as you said, even though N8N does simplify a lot of this agentic flow and the stuff that you were building for this project, um you still have to have some sort of technical aptitude to be able to do it. And so you have to spend time doing it. And I think that's the gap that a lot of people miss when they talk about using these tools to just suddenly build something out of nothing. And that's not even to say we're talking about whatever security vulnerabilities and stuff that you need to know about or at least be thoughtful enough about to factor in as you're building. So this is really cool. So with this N8N, now forgive me, I don't know the terminology for N8N. Is it called like a workflow, a pipeline?

Asif: 17:56

Yeah, yeah. It it's kind of low, and you call you can call it as a pipeline too, because you connect different kinds of system to build a big system. I can define I can define like uh integrating different kinds of systems to build a big ecosystem.

Julian: 18:12

Yep, that's cool. It it reminds me of sorry, I was gonna say it reminds me of like a giant API where you just plug things into it. Yeah, sorry, keep going. Yes.

Asif: 18:22

Uh I just find here because in the real life, for the work you have done, you should have a value for that. Fine. If I've done workflow, but it was not deployed elsewhere. It has no value outside the market. It has at least it doesn't have what has a value in front of a recruiter who is interviewing you and when we talk about the interviews thing. So work should mean something that should work for the real world. So that makes think about Night and more seriously when I was working through I have built so many ML models but haven't deployed into many. Because N80 makes me makes it easier to connect the my ML models with different kind of websites or the data, so it could work who wants that where it is necessary or something. Yeah. So yeah, I I I never stopped because my skills in front of recruiter because the work should mean something to them. Yeah, this is the main part.

Julian: 19:32

That's that's a really important take and and very interesting. So with in you know, we've talked about this before, but with Pie Bytes, with our coaching, you know, we are constantly having this message that you should build to learn, right? That's one thing. But we also say you should build to promote yourself, right? That when you build a project, when you build an application, something like what you've built, you can highlight that to recruiters. But one thing that I just haven't considered is exactly what you're saying is well, how useful is that item that you're building? So we always say build something that scratches your own itch, that solves a problem for yourself, right? And that's fine. But if it's something that scratches an itch that solves a problem, that you can then highlight that value to a recruiter, that's even more valuable. So I'm really happy that you said that because you know, recruiters, they don't necessarily understand the technical content in your application. You could show them all the code in the world, but they might say, I have no idea what you're talking about. But if you can say, here's what it does, though, and it solves this problem for this kind of a business or for these kinds of people or this demographic, whatever, that's super powerful. So I really, I really appreciate you sharing that. That's a really good point. Um so with with this N8N pipeline, are you able to? I know this is this is a production app that's being used, and you know, people are paying you for the service. Are you able to share any of how it works for the people listening? Just without giving away the secret, but how how does it work? So you said you know, there's a chat bot at the front end, something what what's the LLM that does the response? Um, and how does it work?

Asif: 21:17

Yeah, this is kind of a technical concept. It works as a rag system. Yeah. And so first we take the data of the customer. Fine. If if you have a five bytes business and you have come to me to build chatbots for me, fine. I'll I'll I'll ask all your company data first, and then I'll convey enough embeddings and and use a different kind of semantic search between them and store it in a 3D vector database. Fine. And also I'll connect that data with LLM. Fine. I'm not taking your data and Sending everything into the LLM and access the natural language between them. Fine. It just adds hi, how are you, or something in the natural language with the data you gave me. So we can talk the data privacy part over there. Fine. Because I take your data and I have to be responsible for that. So I'm not sending everything into the LLM. Because yeah, the big models train if data to that, the big models again train on them and your data becomes private. So yeah. I created a knowledge base with and just take the natural language from the LLM and give some inputs to the LLM. Fine. You are an agent. You have to behave like this. In a real world example, if you talk, if you talk to a receptionist in the model, you instruct them, you have to talk like this. You have to make them work like this. And you have to assign different tasks to the different kinds of workers in the motel. The agent, what it does, it takes the data from the knowledge base and then it acts it acts as an owner to a substitute between the data and the customer who is asking a relevant question. When this really when this when the person who is using the EUI restaurant, he just asks, when does a restaurant open? On Monday. Because people are nowadays to lace Google maps and see yeah at what time and at what day and at and what time it is the restaurant or anything. The vision is open until. So he just asks, what are you open in timings? And this chat board, what it does, it just takes uh if you've given me, it has a knowledge, this restaurant opens at 7 a.m. in the Mondays. It just recognizes that and adds the knowledge user tone. If we ask anything like politely, hey, can you just give me uh the information about the opening timings of the it just says, Yes, sir, here the timings are. If you just ask in the way, hi, how are you? Just give me the timings. It just reacts to that saying the opening timings are just 7 a.m. And we just have to train the LLM to work like that. This way you can develop a RAG model, like an AI agent which works with the concept called RAG to the different kinds of business, because these days there are the there are very good players in this market about this building AI agents because it's become a no-code thing, and everyone can build it. But not everyone can build a reliable, reliable, and scalable system could definitely be helpful for the big enterprise.

Advert: 24:45

So this comes from learning. Just a quick break from the episode. Stuck at the beginner or intermediate level and wondering what it actually takes to become a senior developer. It's not just more years of experience, it's owning projects end-to-end, making solid design decisions, writing clean tested code, and communicating like a pro. That's exactly what we help people with inside the Pybytes developer mindset program. You build real-world apps, get weekly code reviews, and sharpen your developer mindset that will lead to promotions. If you're serious about stepping up to a senior dev, check out the link in the description and apply right there. Now back to the episode.

Julian: 25:26

No, I like that. And I like the concept you're getting to. Anyone can do it, but not everyone can do it well. Yeah. I I like that. That's that's really cool. I I won't dive. I know you're I can I know that you're this is proprietary stuff you've made, so I don't want to push too far. We can't go into the more technical stuff there for you know uh privacy reasons and everything. But that's really interesting. And I'm glad that, you know, right off the bat you're thinking about privacy, right? Because that's one of the biggest concerns as soon as you start talking about these models that companies can deploy with their own data, right? So you you gave Pybytes as a loose example. If I was to let an LLM loose on our internal documentation and everything, how do what are the guarantees that that data is private, right? And then you think that I have customers, so I have their data, and is that going into the LLM, right? So there's this this flow on effect, and you have to be careful of that. So it's really cool that you're thinking about it and and protecting data right from the start. That's really cool. Um, so you you touched on this. Sorry, you you got what we can say.

Asif: 26:31

Yeah, I can touch down a project I'm working now, which is related to this kind of thing. Should I go on with that?

Julian: 26:40

Yeah, sure, please. Yeah, more projects the the better. Go for it.

Asif: 26:44

Yeah. Here comes the thinking working with all this AI agentic stuff. I was always calling the LLM, which is like Gemini, Deep Sig, and Chat GPT models, uh act as a natural language. But when I went to when I went one to one of the enterprise and he said, yes, to all the LLMs, what's the guarantee level you could give me that the data is not passed into the LLM? So I was I I was like blank. Fine, that might be true because I'm connecting with an API calls with an open AI model, which base I developed for an organization. That might be a data lake.

Speaker 1: 27:25

Yeah.

Asif: 27:26

When you go to the good enterprise, the big enterprise form people ask, people ask for that, definitely. So I went blank for like 10 days and have searching all the GitHub stuff and all the YouTube stuff. Fine. What to do that? Like how how data is stored in this? No, how how how reliable it is and how securely it is sending the transmitting the data between the LLM endpoint and the database endpoint. Fine. Here comes a solution. When you have your own server, when you have your own MOE, you have your own database, why to bring your why to bring an outside LLM? You can bring your own LLM. You can find training training that requires a lot of GPUs and CPUs, but you can download them from Hugging Face. Yeah. That's a top-notch technology right now. Downloading the models, yeah. You can you can also that's really a big boom. You can where you can download the models, and fine. You and you can there are the project is about the document passing system where it reads all the documents. Fine. Here, here the solution is you can there are many LLMs which work on this passing data set. You can take one of your LLM and paste that in your server, and you can also college base inside your server, and from then you can also have the connection between the endpoints and the chat, and you have a chatbot or where you want to use that LLM, and you can connect them with your server, not without an outside server. This is a own server, you have your own LLM, you have your own database. They solve my equation, and from then I've asking the business for the servers, and I'll I'll I'll run an LLM for that for how the organization wants it to perform. Yeah, that's a good business idea these days, and there's a big boom in this technology and in this business area field developing an organization without touching and a cloud cloud LLMs, like chat, Chat GPT GM.

Julian: 29:44

Yeah, I think that's a that's a really that's a really cool use case of almost scratching your own itch, right? Because you talked about OLAMA and hugging face and things like that. And I know people in our community who are using Olama internally at home, right? And one of my friends who who'll probably be listening to this episode when it comes out, uh, he was tinkering with Olama at home running on his laptop, right? And I know people who are running these things at home because they don't trust pushing information into Gemini or you know, OpenAI or whatever. Uh and so being able to keep it completely local with no access to the internet, other than you updating it every now and then, um, is super powerful. But what's nice about what you're saying is that you then took that concept because I know a lot of people who just want to do it for home so that they have access to an LLM that's you know as as private as you can get, as opposed to using something online. But um to then take that and say, well, now that I've created this private instance of an LLM using Olama or Hugging Face and so on, you can you then go, okay, how do I monetize that? How do I turn this into a business? How do I keep that privacy first um sentiment um in a business idea? And I think that's really cool. I I like the way your your mind works with these things, it's very cool.

Asif: 31:15

Yeah. Um about I was I was the I was failing every time while building this. But the main no the main you'll get a recognization or you'll get up to the top-notch knowledge person when you pass all these screens. Failure, failure comes every time. When when I first tried to deploy an LLM into my system, my system almost crashed because I downloaded one TB version into my yeah 512 GP SST Mac. Uh uh it it I just got destroyed. Like, I don't know about this. Fine. The next time I download a 256 GB version, which works well on my Mac. Fine. Everything comes from the field with the one TB version, and you came to the 256 GB version. And in future, if you get a GPU, you'll definitely go to that version because it you know you have failed over and it how it how it's gonna work. Yeah. When you fail, you can know how how you can gonna make that work. And that's the ego you should have, actually.

Julian: 32:17

Yeah. I like this this is why you're a good teacher. Because I like that. If you this is what you're telling your students, that's awesome. Um, so actually, and that leads me to the question I wanted to ask is as I'm listening to all of this. How did you learn it all? And and I know it's I think you've answered that question, right, through experimentation. But is there anything and so here's a tip, actually, here's here's a way we can phrase this for everyone listening. There's so many people who won't even get started with this. They won't even get started with uh looking at whether it's machine learning, data science, but also just just agentic stuff, you building stuff with AI agents and LLMs, they won't get started because they don't know where to start or because it's um conceptually complex. So when people are daunted and kind of scared or even just not willing to put in that effort to start, what kind of advice do you have based on your experience learning all of this now? What have you learned from that experience that you can share with people to maybe inspire them?

Asif: 33:20

Yeah, yeah. I have so many uh or a lot of my friends who are working on ML or AI domains. For that, you initially have everything for every to know Python. Fine. Where can you get the Python from? You can get from different platforms. There are many repetitive platforms in the market. Uh, I can say such as like lead code and the pie bytes, where you can grind your knowledge on. Fine, everyone, every uh people stopping early because while right uh because people get to know what's a fur loop and while loop initially. They see it as like a big astroscience, and they just stop it because when they get habituated for doing that fur loops and while loop, it gets better on daily progress. When you grind on something, when we focus on something, we should move on. Like there's a saying from my father, practice makes a man perfect. Yeah. So I followed this principle, and I also suggested to so many of my students and my friends to not stop thinking it is an astroscience or something. Fine. In the first day of learning, fine. Take me take an example of me. In the first day of learning what is a furloop, I just I just put my bench and I was like, what this furloop for and why this whiloop for why I am learning this? Yeah, and it all began like that. There's there should be a little bit of ego everyone should have in their ego in learning. Because to learn more, so you can be a more pro efficient if you learn more. You can grind on something for like ages, for like years, and become process at a certain point when we don't stop at an early career, early early starting of Python. And for machine learning and everything, need to not touch data structures and algorithms concept in that, which which everyone feels it as a superior structure. I do feel it, but everything everything comes on a grind. So uh when you want to be a machine learning engineer, you definitely definitely have to be focused on the different concepts like numpy, panda, cycitals, and tons of flow models for the deep learning and stuff and all. You definitely need like you should have a basic knowledge about data structures, complete knowledge. So don't stop if you see data structures while becoming a machine learning while you want to become an ML or AI engineer or research. Don't stop with Python if you see data structures. Go beyond that because data structures only apply to the different kinds of soft fields or something. It definitely applies to this machine learning or AI, but not completely. You said you just have to know the basics among that plate structure, not a complete lead code kind of thing, which every fan company is to ask for. Yeah, so don't stop. Know what you want. Like fine, if you want to develop an ML pipeline, you just have to need the basics of Python, not the basics of my knowledge, about the Python. I'm not talking about the data structures over here. So it's just about Numpy, Panda, Cycit Learn models or something. If you want to be but there are no good, there are no good tools in the market. Yeah, yeah. You you can call it as a wipe coding.

Julian: 36:44

Yep, yeah. Okay.

Asif: 36:46

So so never stop early because you saw something destructive in your path. So move forward and know wait what actually needs for my end success. And learn until that and move on. If you if if if something comes in your way, try different kinds of resources, try different kinds of platforms, try different kinds of professors, lecturers, mentors, and you definitely have to, we'll definitely pass through them.

Speaker 1: 37:13

Yeah, nothing.

Asif: 37:14

Nothing is a neuroscience. Nothing is a neuroscience.

Julian: 37:19

No, this look, this is cool, and it's kind of really reassuring and comforting to hear the same the same concept of learning that not only that we push, but that we all know to be true about learning anything, you know, the guitar or learning to drive a car or learning any skill. It's the same thing. It's push through it. It's going to be challenging at certain points. So just be persistent and consistent, keep at it. And then when you get stuck, go and seek help. You know, go and talk. You said go and talk to a lecturer or professor or whoever is senior around you that you can do it. And if you don't have seniors, here's a nice plug for the Pie Bites community. Come and hang out with us. And if you don't have anyone around you, you can ask. You could ask in our community, you can join our coaching, that kind of thing. But that's the point, right? Is that you learn, you get stuck, you push through it. If you can't, you seek support, and you just keep doing that until you get to that point. And I think the other uh the point of success that you're after, but the other thing you said that I think is worth repeating and drilling in is just that you know you learn what you need to learn, right? You you know what it is, you look at that end goal, and this is something we talk about a lot, is that kind of um working backwards concept of going, I need to get here. So to get there, all I need to learn is X, Y, Z. I don't need to learn A, B, C, D, E, F, G, or everything to get there. Just XYZ will get me there. Because you sometimes don't need to learn everything and every single thing to get there. You learn it when you need it. And then when you get stuck or something breaks, that might require learning a new skill, such as deployment or whatever, right? So I think this is really cool. And just to wrap this up, so you all of this, all of this learning, all this stuff came from just having an idea and then just experimenting to you could see that idea through. Is that correct?

Asif: 39:23

Yeah, it's definitely correct. You have to you have to grind, fail, grind, fail, and at some point of your life, at some point of pass. And that's uh that's a mindset you we should have while learning, while coming into this ML or AI thing on in the Python. Nothing is hard. So if you just give a try, if you just give it try two times, if you just give it try ten times, you definitely learn it. So it should be an unstoppable journey in the education field.

Julian: 39:55

That's cool. I love it, man. Lifetime learning. That's that's very cool.

Asif: 39:59

Yeah. So but not every time learning after the ages, too, but at some until some point, until you achieve the greater heights in your life.

Julian: 40:10

Then then you learn something different. Then there's other stuff to learn. I'm getting to that point. Uh there's stuff beyond Python that I want to learn. So that's cool. Um, all right. So as we wrap this up, um, Asif, you know, one, thank you for all for all the insight and for sharing all of this. It's very inspiring to me. And I hope everyone listening, you know, feels that as well. That, you know, just if you have an idea, you can just there's so much out there now that you can just uh take it, take advantage of the technology out there to build cool things, but you do need to still slog it out and still push through it because it is going to be challenging when you get to those upper echelons of what you're building. But for you, you know, if I was to ask you, this is a challenging question for some people. Um, what is next for you? I mean, and I ask you this because you're so relatively young, and I don't mean to harp on about your age or anything, but just there's so much life ahead of you, and you've already accomplished this much with building these apps for these, you know, booking agencies, hotels, and things like that. Um, what is next? What what do you think you're gonna do after this?

Asif: 41:13

Uh in the in my short term, I just want to work for like a fan company, like a seminal engineer or a researcher or a scientist. I'm gonna more knowledge. And you know, when you work for the good big enterprises, you know what the real problems are outside. And in my long term, I just want to habituate my like building an agentic AI systems as a business in my long term. Let's see, and let it happen in like a six to seven years for that. I'm just planning to, yeah, after getting all this knowledge and you know, from the big companies, from the like exposing to the outside world and networking and meeting the people like you. I just got to know how like I'm gonna engage. Fine, I just want to have more knowledge how the outside world works and what it really. So, according to that, I'll just adapt to that changes, and in the long term, I'll just propose my own company, maybe something for that.

Julian: 42:22

Yeah, great idea. I mean, of course, get out there, get some experience, and then start your own company. That's awesome. I I think that's cool.

Asif: 42:30

Um, and and also I just want to mention the Pybytes platform here. When you just first introduced when you when I just first thought what's Pybytes platform means, fine. Julian introduced me to this Pybytes platform and he's a co-f I just have to look how it is, and I just went through all this, you know, the discussion parts and the coding parts. For real, I just felt lead code was old. Not I'm not talking about the different kind of coding structures, but lead uh the platform such as lead code doesn't allow you to learn and code, it just asks you to code. PyBytes platform makes it different, it makes you learn and practice. Learn and practice is a perfect combination other than just practicing. And you could find this, you can find these two things in the same platform, which is like PyBytes. Open discussions over there and the way the students interact between them. I wish back in like when I was learning Python, I uh like I got this Pybytes on the uh cloud. Thanks, man.

Julian: 43:41

I wish so. Oh, that's really nice of you to say that that's very nice. I was not expecting that. So thank you. I I I do appreciate that. Uh and it means a lot coming from someone like you who is is quite established and you teach, you know, you're a teaching assistant and advanced Python. So you know your stuff, and for you to look at it and say, I oh I like that comment, man. That that's made my day. Lead code fe feels old. That's wonderful. Um I appreciate it.

Asif: 44:08

Um about learning and practicing.

Julian: 44:12

No, I love it. And that's that's it. So you just gotta keep repeating that, and uh and you'll get there. So the the last thing I always ask people on this podcast is if um, and not that you're not busy or anything, but are you reading anything at the moment? Any books?

Asif: 44:29

Yeah, I'm I completely depend upon the knowledge of video streams and different platforms. But in order to develop my character base, fine, how much knowledge you have, you just have to there has to be some basics in your character, like a this something called as discipline in your life to achieve all this I have. For that, I read a book called The Subble Art of Not Giving, and that made my like it's not about it's not about all the negative impressions, it comes about with failures, and I like that. And I'm developing like that, I like that, and I recommend to everyone it supports you to develop a discipline upon the negative things.

Julian: 45:18

Yeah, is this the saddle art of not giving a by by um Mark Manson? Is that the one? Yeah, the book by Mark Manson. Yeah, yeah. I think um I don't know, I don't think I have that on the bookshelf because I don't want the kids seeing it just because of the name. But no, I um I I that is on my reading list. I just I never got around to it. I've watched some of his YouTube videos, so I kind of think I get the gist of it. But that's that's cool, that's a good recommendation. Thanks, man.

Asif: 45:48

And also and also an idea for someone who is developing in their life because I got from someone special in my life to yeah, to get that. Yeah, you'll learn discipline like that. Not because you don't know, but you want you wanna know that.

Julian: 46:04

That's cool. No, no, I appreciate it, man. That's that's very nice. Um, I'm not reading anything new since last week on the new podcast. But I will say, um I got a book, I can't remember, I think it's called Hope in Action. Let me see, one second. Yeah, it's it's Hope in Action. It's by um oh my gosh, what's I think it's Anna Maron. Uh anyway, she used to be the Prime Minister of Finland. I think I should really have done my check before I um before I started talking about this.

Speaker 1: 46:39

Yeah, so yeah.

Julian: 46:42

Um, but she she released this uh amazing book. Um I I saw a an interview with her online as she talked about the release of the book. And so I'm really looking forward to it. I think she was like the youngest prime minister um and you know female prime minister in the history of of the country, something like that. And anyone listening, please correct me. I I know I've got something wrong there. The book is just out of arm's reach, so I can't even confirm this without reaching for it and having an awkward pause here. But uh anyway, I'll have all that in the the comments along with your book as well, uhsif. So um in the in the show notes and we'll we'll get to that. So look, we'll we'll wrap it up here. Um Asif, you know, thank you again. But just quickly before we go, where can people people will want to connect with you, right? So if if people listening to this say, you know, follow a sif follower's journey, connect with him, learn from him. Um, I know you are in our Pie Bytes community now, but where else can people find you?

Asif: 47:39

Yeah, yeah, yeah. People can find me in LinkedIn actually. I'm very active in the LinkedIn. So yeah, I'll just post a link with this.

Julian: 47:47

Yep. Okay. I've got your LinkedIn anyway, so I'll get that as well.

Asif: 47:51

And also, and also up my personal website, I'll also just share the link up and they could look out to my portfolio and all my projects are over there. It might excite them because it's it's better like a Netflix type.

Julian: 48:05

I loved your portfolio website. I showed it to Bob and thought, this is so freaking cool. So uh yeah, please, everyone listening to this, click on the the portfolio link below and you know, tell me it's not next level, it is just such a cool, cool portfolio website. So we'll we'll don't ruin the surprise, we'll make people go check it out. Um all right. Well, that's it. So look, as if thank you so much for joining me. Um, it just what a pleasure chatting with you as always. And uh, you know, a lot of the stuff you mentioned today we hadn't discussed, and I didn't know a lot of the detail and the technical detail there. So I I appreciate it. Um, thank you for being so forthcoming with what you're working on and sharing. I know there's a fine balance between sharing and keeping things you know proprietary and quiet, but I appreciate it. Any any last words for everyone before we go?

Asif: 48:56

Oh never stop because of your failures and try to take your ego out for success.

Julian: 49:06

I like it. I like it. That's it cool. Take your ego out for success. Yeah, that's it, man. This is it's a lot of it is mindset. I I appreciate it. That is that is very, very cool. Um all right. Well, look, uh stiff, thank you. What a pleasure. Um, you enjoy life over up there in Arizona. Um and I can't wait to see what you do next. So thank you for that. And when you launch the next thing, you'll have to come back on and join us and and share about that as well. Um thank you for inviting us. Oh no, my pleasure, man. Happy to have you here. Anyone in our community, just so so glad to have people here doing cool stuff, and that's one of the main reasons I wanted to bring you on here as well.

Asif: 49:48

So and and that's the most important thing I like about Pie Bytes because people what they are doing, and the the people who are not doing can take inspiration from them, yeah, and work along. Yep. I'm I'm happy to be here with you. Yeah, and thanks for this opportunity to be with you on in your podcast and a pie bytes. There's uh I think there's gonna be a big boom in the coming years because I see all these advancements these days, yeah, big boom, and everyone follow follow the PyBytes platform to learn Python and to engage in different kinds of open discussions and real life projects and expose to the real world people. You can't get that with your fine.

Julian: 50:36

That's awesome, man. Thank you. There you can be our marketing person. I I appreciate that. So, yeah, please join the community, use the platform. Uh sif's all over it is awesome. So, thank you, man. I appreciate it. Everyone, thank you for listening. As always, thank you for tuning in. Uh, please make sure you do like, subscribe, share, review, whatever it is. Um, and we appreciate you listening every week. Until then, we'll see you on the next episode. Cheers, bye. Hey everyone, thanks for tuning into the Pie Bytes podcast. I really hope you enjoyed it. A quick message from me and Bob before you go to get the most out of your experience with Pie Bytes, including learning more Python, engaging with other developers, learning about our guests, discussing these podcast episodes, and much, much more. Please join our community at pybytes.circle.so. The link is on the screen if you're watching this on YouTube, and it's in the show notes for everyone else. When you join, make sure you introduce yourself, engage with myself and Bob, and the many other developers in the community. It's one of the greatest things you can do to expand your knowledge and reach and network as a Python developer. We'll see you in the next episode, and we will see you in the community.