Conversations on Applied AI
Welcome to the Conversations on Applied AI Podcast where Justin Grammens 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 AppliedAI.MN. Enjoy!
Conversations on Applied AI
Dishant Gandhi - Building a Community for Chatbot Developers
The conversation this week is with Dishant Gandhi. Dishant is an AI/ML Engineer at Hoory, who reads an AI-powered Customer Support Assistant that helps your business automate customer communications and provides 24/7 support with zero delays. Dishant is also building a community for chatbot developers and learners. He has a YouTube channel under the name Droid City where he creates videos on chatbot development and Android app development. Finally, he is a RASA Superhero, which is a distinction given to the most contributing members of the RASA open-source community.
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
- Hoory
- Droid City Youtube channel
- RASA open source community
- Turing test
- Machine Learning Specialization on Coursera
- How to Win Friends & Influence People by Dale Carnegie
- Talk Like TED by Carmine Gallo
- Hands-On Machine Learning by Aurélien Géron
- Natural language processing
Enjoy!
Your host,
Justin Grammens
Justin Grammens 0:00
Greetings Applied AI Podcast listeners. This is Justin Grammens, your host of the Conversations on Applied AI Podcast. Just dropping in to let you know about a very special event we have coming up on Friday, May 12. It's the spring 2023 applied AI conference. You can learn more by going to AppliedAIConf.com. This full day in person conference is the only and largest artificial intelligence conference held in the upper Midwest. It will be in Minneapolis, Minnesota on May 12, we will have more than 20 speakers with two tracks covering everything from Ai, business applications, chat GPT, computer vision, machine learning and so much more for being a listener to this podcast, use the promo code podcast when purchasing your ticket for a 50% discount. So here are the details go to AppliedAIConf.com. That's AppliedAIConf.com, to see the full schedule and register for the only and largest artificial intelligence conference in the Upper Midwest on May 12. And don't forget to use a promo code of podcast when checking out to receive a 50% discount. We look forward to seeing you there. And thank you so much for listening. And now on with this episode.
Dishant Gandhi 1:08
I say the framework for building assistants on top of it, you should first of all have Python knowledge and basic programming language knowledge so that you can build an assistant on top of it. Because this is a Python library, which you can use. And you can create different kinds of I won't say a specific because it is just a framework where you can have you can use bird, you can do die classifier on top of it. And you can directly train whatever questions you have. And once we have added questions, answers, and whatever API calls or database connection you want to have, you have connected you can bring in the model, and you can chat with that particular assistant.
AI Announcer 1:50
Welcome to the conversations on Applied AI podcast where Justin Grammens 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 2:20
Welcome everyone to the conversations on applied AI Podcast. Today in the program we have Dishant Gandhi. Deshant is an AI ML Engineer at Hoory, who reads an AI powered Customer Support Assistant that helps your business automate customer communications and provides 24/7 support with zero delays. Shawn is also building a community for chatbot developers and learners. He has a YouTube channel under the name Droid City where he creates videos on chatbot development and Android app development and have checked out his channel and watched a number of these extremely useful videos. They're really great. Finally, he is a RASA Superhero, which is a distinction given to the most contributing members of the RASA open source community. So thank you so much to shot for being on the program today.
Dishant Gandhi 3:00
Thank you so much, Justin to invite me in some easy podcast. So really excited, what kinds of questions and what we'll be talking about AI and Machine Learning.
Justin Grammens 3:11
Great. Yeah, well, we're super excited to have you because people are always looking to learn new things in our community. And every one of these podcasts, somebody learns something. And in fact, you oftentimes it's I learned something doing these. So I'm really, really thrilled to have you take some time today. And you know, enlighten us on the work that you've been doing. So I gave a little bio little brief introduction there. But maybe you could give us a little more background on kind of how you yourself got into AI in your career.
Dishant Gandhi 3:35
I like to describe myself, as I'm really passionate about creating the virtual assistant and bending assistants, which really help solve the problems of the real world, I myself have faced issue while connecting with the company, I go there, and I have to talk with the people. And it's really time consuming for everybody. Like you have to be in queue for just a simple question. Like, how can I book this, I'm not able to book this particular ticket for my flight. And you've just had to wait in the queue before days. But now it's then like you call or chat with the assistant on WhatsApp or as the messenger wherever their site is, I can just answer it. So this is how I like to restrain myself and also, as you described, like, I also have a YouTube channel, which I created because I was seeing at the developers who are building assistant or are in ML or AI, they are really having trouble learning about this if they are beginners. I myself had troubles learning AI and ML before but nowadays there are quite a lot of tutorials and I'm also trying to contribute as much as I can in this feed so that everybody can do this themselves.
Justin Grammens 4:52
Great. No, that's awesome. So I think I saw you know, you went through some coursework I guess on like maybe Coursera and and before that, were you You just writing software or you some of these, you know, your software engineer and he said, Well, wow, this AI seems like a really interesting place for me to get into to contribute. You kind of made that mental shift and kind of went back and started learning this stuff.
Dishant Gandhi 5:12
No, actually, it was not like that. So since childhood, I really had the craze about AI and machine learning, I really wanted to go into robotics. But that was not the part of my job. So I really liked coding part, instead of building some electronic machineries like Iron Man did. So I'm not bad. But yeah, the motivation from where I, I thought this was clearly from Avengers then signed such to paid and, and all those cartoons, which, which really helped, like Power Rangers was also some kind of AI they used in there, like watching those cartoons and shows it really helped me. So whenever there was some some kind of show is really I was watching it properly, like what they did, and how it was working. There was also a show called as brain games, I remember on National Geographic Channel, they also showed some kind of technology, I don't remember because it was like, many years back. So yeah, it really helped me to get into this. And in the degree college when I when I was in my degree College. At that moment, I learned about Alan Turing test and a book about TI, it was just a theory book not hands on well, but they really explained about Turing test like how it is done, I guess you might also be knowing about this during test, where there was a person who will sit they will give some responses, he will also write a response and the machine will also write a response for a particular test to conduct and the Jacob the machine is working properly. And it is actually AI. So is the responses are similar or mad in some way. It is known as it is successful and idle, it was published in 1915. So yeah, it was really amazing. Like how at that moment, AI was defined, and we still use it like, when we are creating a chatbot or we are talking with an AI, we try to check if the questions which we are it is really able to answer it or not in in a really conversational manner. Our children is
Justin Grammens 7:18
one of the hot things these days, which everyone's talking about is chat GPT.
Dishant Gandhi 7:24
Yeah, definitely, it is a really amazing topic.
Justin Grammens 7:27
I use it pretty much every day in one shape or form, because I'm generating and writing a lot of content myself. And I'm not using it to generate a lot of things, but also to sort of actually make what I write better, and also give me better ideas on things that I hadn't even thought of, and then I'll grab pieces out and be like, Oh, I should write about this. But yeah, the Turing test is interesting, because it's giving me responses that it could be human to me on the other side, I mean, it's almost like and that's kind of what the Turing test is, is you wouldn't have any idea if it's a computer, or you're talking with somebody else. Where do you think we are with regards to the evolution of chat? Just I guess, chatting about chat GPT? But what are your since you've kind of you're deep into this space? How do you think we're doing? And are we doing really, really poorly, we're doing better? How do you see this going?
Dishant Gandhi 8:09
I would say we are growing really better. And this particular area, because I remember when I really started my degree in BSC it which is Bachelor of Science in Information Technology. At that moment, Google announced that they will be creating an AI or they will be coming up with something which will answer your questions in a way that you won't know that it is a person, or you have just started to chat with someone at dark moment, the 4g in India was just launched. And we actually went in the college time here using the Wi Fi, open Wi Fi, which is available by Google on the on the railway stations. So it really helps to know about this all things and we use Wi Fi for doing this all stuff they in this all experiments on programming. So yeah, we are really growing much better at this rate. And I would say they are experimenting few things, which it's really good that they are not sharing everything with us by Yeah, it would be really great if their developers who are really involved in this field would get to know about the experiments which are going on. Yeah, yeah.
Justin Grammens 9:20
I mean, it seems like of course, it's only as good as the data that it's been fed. And there's lots of questions around the data being bias, and it's not being up to date, because like Chechi T is still a number of years out. Going back to what you guys do at worry, you know, you guys are doing customer support. But how do you guys feed the bot that hey, I'm assuming a hotel is going to have different information that's going to be different from from an from an auto store, you know, with regards to customer support. I'm just curious how is every implementation that you guys do very unique and very time consuming for you to build that out?
Dishant Gandhi 9:54
No, I would say is it's time consuming, but in reality, our product hurry is a concept where you can build any kind of Chatbot. Or I won't say chatbot, but a virtual assistant, which will help your business grow and solve your problems like it has live agent abilities. But you don't need any kind of coding knowledge or programming knowledge to build the assistants on our platform, you can just add the questions I add your training data, or whatever you would like to say, variations is also a part of this. And you can add answers or whatever response you would like to have for further assistance questions. And that's it, you're trained the assistant, you integrate with the website or wherever we like to integrate it. And you can talk with the assistant and your customers can talk with that session, solve the problems which you have or the questions they have with that. So it's really simple, because each and every business each and every startup does not have fundings, or it does not have capabilities to hire a programmer or developer to help them create the assistant because nowadays, there are multiple companies are providing artificial intelligence, but what happens is they really have limitations, with what user can ask board can actually ask and response to the user. So it really helps in a conversation manner. They can communicate with the assistant and get the response in a conversation manner it it should not be like, Okay, so this is the following step you have to do. This is the step number one, step number two, we have really got acquainted, I'd say in the robotic manner of the assistant, but what is some assistant is able to answer your question and very conversation, you won't be able to know if there is a particular person behind it or an assistant. So yeah, we are currently building that kind of system.
Justin Grammens 11:53
Neat. And how long has the company been around?
Dishant Gandhi 11:55
I don't know, a specific year. But I can say there are probably two to three years, or more than that.
Justin Grammens 12:03
It's such a hot space. And there, I've seen a lot of companies that are sort of getting into this. And I could see as a startup, you're right, you not only have the programmers or developers to do it, but you can't even pay a support staff. I mean, you're, you're busy. Oftentimes just trying to sell the product or sell your product, right. And so to have a virtual assistant, that's there 24/7 That can interact with people that have problems, while using your product or need support. To put that off on an AIS is great.
Dishant Gandhi 12:29
Yeah, definitely. Because nowadays, the internet has grown so much that if you are selling your product in India, or in a particular state, someone can call you and get it from any other state or any other part of the world. So to be available 24 by seven, it's really tough job. So the system can solve this problems for you, even the assistant can get orders for yourself, and you can then ship it out afterwards, once your time zone is available.
Justin Grammens 12:59
Yeah, I was gonna ask you about that, you know, it's one thing just to put a frequently asked questions, you know, into a bot, right? Where it's just like somebody asked a question, and then they get a response back. But it sounds like your system. Can I guess a couple questions here is is Yeah, can they place orders, I guess through and it seems like SMS, you know, text messages is also a way that people I see a lot of chatbots and like interacting, how, what's the interface? I guess? So I guess there's two questions there. And number one is, is maybe what's the capabilities? You know, is it just back and forth chatting? Or are there other integrations or API's into the back end? That's probably my first question.
Dishant Gandhi 13:32
The assistant which you build is completely based on the training data which have faded. So if you fed it in a really good manner, like you have enough training data and responses, then the assistant will understand it properly. And the pipeline which we are using, it's really good. So that even if you have questions, like even if you like added few questions, it will be able to answer in a more conversation manner. So you don't have to worry about it.
Justin Grammens 14:00
Gotcha. Gotcha. Yeah. And then I guess the response could be hey, here's a link to our website, go ahead and place the order, you know, and I wasn't sure if you guys have like text message capability that people can text with with your with a chatbot or is that something that they would have to set up but then feed your model and tours?
Dishant Gandhi 14:15
No, they can integrate the widget itself in their website and any consumer or any one who is visiting their website, they can ask the questions and also they can communicate with the assistant. So yeah, also we have live support functionality, where if the assistant is not able to answer it will ask for few things, few few things from the user view information. And then once it is captured, it will wait for the support agent. If they have enabled them to care. They can directly answer the questions of the consumer. So yeah,
Justin Grammens 14:53
nice. Yeah. So you, you can kind of have a human in the loop, I guess is what they say. Sometimes you put in human into it. Where do you see this? going in the future? I mean, I mean, obviously, I guess it's just going to get better and better. But, you know, you've been in this chat space. I mean, you have a YouTube channel all about it. And we'll put links to that, along with, you know, links, links off to Huri in the liner notes as well. But where do you see this going in the next, you know, five years or so
Dishant Gandhi 15:17
That the only thing about the chat space going out, say, instead of chat, people might be going hands free, like using directly voice because everything is going wireless. And nowadays, I've seen mobiles, mobile smart smartphones, does not have aux cable. Now aux cable even so there is no wire, you just click on your headphones. And that's it. You can directly call a CD or a virtual assistant, any kind of joint which is integrated, and you can chat with them. So yeah, I'd see that each and every space will have voice assistant and also a wake up call, like, Hey, Siri, can you do this for me? Or hey, Google, or this is already present, but you're not using extensively. But once it is done it when it would be really good to have
Justin Grammens 16:08
Fair enough. Yeah, for sure. I mean, I in fact, as I was walking in today, I wanted to open an app on my phone, we have a security app that we use to get into our office, I just used voice to bring it up. I said, Okay, Google and open this particular app. And it did. And, you know, it only saved me a couple of seconds. But I did have to flick through all these apps and try and find the thing, just so voice control is awesome. And I also think as I'm getting older, my eyesight is actually not as good as it used to be. And so voice it's way easier for me just to dictate something, it just even as simple as a text message or whatever it is, then sit there on keyboard to go back and forth. So voices voice is definitely huge. I just don't think it's used enough. And I think you're right, I think I'm finding it more and more myself realizing Hi, I should I should just say this. And with chat getting better, you feel like it's, it's gonna be more and more integrated into our lives. So if I want to schedule something, like a maintenance thing for my car, for example, I'll just be able to say, okay, gee, schedule this at this particular place, and we're going to become more and more not have to actually pick up the phone and talk to a human, you can just talk to a chatbot
Dishant Gandhi 17:09
chatbot to answer each and everything, it's quite difficult. So, I am quite not often with with AI D each and everything for us. Because then what we will be doing so as as we sit here legged, they will be doing the chores for us. Will they washing our clothes for us? Like, yeah, there is a machine for it? Definitely. But how much you will get lazy for them? Like, will it be a strength or a weakness? I don't know how it will be like being lazy.
Justin Grammens 17:40
That's a great, great thing I hadn't brought up yet. But I usually do bring it up with people is kind of like, yeah, how is this going to impact the future of work? You know, for people? Do you view it as a positive as a negative? Are you kind of on the fence? You're not really exactly sure you were kind of touching on that when you answered that last question. But do you see things where you're afraid of it getting too smart? Or do you see ultimately humans still being able to contribute to the society or you just we just don't know yet?
Dishant Gandhi 18:04
I would say the third option is good for me, because I'm not that kind of person who thinks he would be quite harmful like, as it was in Terminator movies. So yeah, like that, but not assets. But I'd say we should not be fully dependent on it. Madea at some, at some point, yes, we can use them. It's not a using ourselves, but we should use them as our help or guidance. Purpose. Because right now, I also heard many news about Apple Watch, where it really helped that particular person to get aid because like that person had a car crash. The Apple was directly called an ambulance and they are emergency con context directly. And they reached it. He got here so yeah, this way it is really helpful by the way like ACD do this for me. Hey, Siri, do this for me. So it really makes it being lazy just sitting around and the Yeah, let the AI do this stuff. Open the door for me and all these things. It really makes you obsessed on. Well, bass at really gets triggered. And now in a body. So yeah, yeah. Oh, sure.
Justin Grammens 19:20
Sure. There's a whole kind of movie around that I forget what it's called again, but it's it's, it's a good 1015 years old, but it sort of plays on that idea. I think it's Wally. Is it Wally? Yeah, it's just this idea that the world sort of turns into all we do is sit in chairs and the chairs fly around. And while all we do is eat and watch TV, so kind of sad actually. But it's but that's the undertone of the movie is that it can definitely go too far. You know, I've heard the term chatbot versus assistant and I was just wondering, I mean, in your mind, are they are they the same thing? They just kind of two different words for the same thing or is is there would you define them differently in any
Dishant Gandhi 19:56
I Chile the definition is quite might be different. entitled, I'm not sure about it. I haven't read the definition on Google, like Chatbot. In any anyone's mind, it would come up as like, it's just we're chatting bubbles, but assistant is like, assisting you. In any any case or scenario you you'll need assistant for. So yeah, not sure about if little bit sooner or later.
Justin Grammens 20:23
Sure, it feels like a system is more broad, right? It's sort of helping you do anything. And then Chatbot is really just like, the way that you communicate with it, where people think about it's typing, but doesn't need to be typing. I mean, so that's another side of it. I'm just, I'm thinking this about this audio thing, you know, are you seeing, I guess I could see in the future where it doesn't actually write it back on the screen, it would speak it back to you, for example, you could actually carry on an audio conversation, the bot,
Dishant Gandhi 20:47
like chatting can happen like on voice or directly like typing. So both the way says that it's a chat, but assistant can also be a human, and also can be a virtual, virtual personal virtual machine, which can give you an assistant, this is how I look at the assistant word, because it it can be anyone, it can be machine, or it can be human. But in a chatbot, it's a word definitely for chatting with the system, actually with the bot. So I'm quite, I'd say, familiar, I quite a lot speak assistant engineer Chatbot. So every time I, I try to say chatbot, it's really sound, we're doing it with a chatbot, it can be our daily recommend to have this as a voice. Because typing could solve few issues. But why the voice we can directly start communicating more and more because I myself, if I want to solve a particular query, or I have a query with the customer care, I would like to call them instead of chatting with the assistant because we have a lot bigger query, then we could type it. So I have this this this, then our assistant would say you can do this this, but no, this is not what I was talking about. I want this kind of solution for this particular purpose. But the system says that you should try this particular method, which will help you to solve your issue. But that's not what I wanted. So yeah, I thought if it would be more of a voice, then it's really great.
Justin Grammens 22:22
Yeah, you got to kind of come to where the customer is, what they're how they're used to interacting. So I wanted to talk a little bit about like, what's the day in the life for an AI ml engineer at your company? Are you are you poring through a lot of data? Are you trying different models? And a follow up question is going to be like, how do people get into this field? Right? If you were starting and just coming out of school now? Yeah, maybe you could give a little bit of color around like kind of like what, what is the day look like somebody in this field,
Dishant Gandhi 22:46
they start with adding the features in the assistant. And yeah, if I start from the beginning of the day, then the get motivated and start the word I would recommend to go gym or go for a swim, do something which which could activate your brain. And so because this is really important while programming, so if this gets frustrated, then even if it tried to solve the error, it didn't take days to solve Well, I can remember there is one error, like it was coming again and again, whatever I try. But I thought, let's take a break. Let's go for a walk, I usually swim or go for a gym or go stargazing. So this all kind of stuff, I go for it. So I do it. And then I came back, I look at the code. This was the issue I was facing. And it really got solved in like 30 minutes with writing the code and doing it. But it took me like one day to think about it. And I got really frustrated about it. So yeah, it really helps if you try to do some activity along with the work, because it really opened the brain. So apart from that, in the day, a day of lies is I won't say we upload the model or download the mod and we train the model those things. But we really start with having a good amount of work done with what we have, like let's say we have this particular concept in our mind, if I just discuss it out and think what would consumer think of it, when it comes to us? How will they react and how it worked for them. So we will think about it. And then once that is done, we start to implement it in a way that it will help each and every consumer or business who are going to use our product. So yeah, it's not just training the model or uploading or adding some kind of data, but we really think of a feature which will really help businesses to move around and use it in a free way.
Justin Grammens 24:52
Gotcha. What programming languages are you mainly working?
Dishant Gandhi 24:55
There are a lot of programming languages see we have and yet so Basically, Python is the most used programming language in AML. That's the one and not to say, we use JavaScript also, for this particular verb. See, there's two other major languages we are using right now.
Justin Grammens 25:16
Gotcha. Yeah. So if you I was going to college, I was coming out of college, I should try and pick up those two languages would definitely give me an edge. Are there other? I mean, obviously, you have your pod, your YouTube channel, where people can go and and learn stuff. Have you seen other interesting sources of knowledge or sources of information that maybe somebody would be able to get for low cost or free in the space to learn things?
Dishant Gandhi 25:40
Definitely, there are really amazing courses available on machine learning AI. I personally would like to recommend Andrew Ng's machine learning specialization, which is on Coursera. You can even audit the course you can learn from it. That's the last of it. If you want the certification, you can pay for it if you want. But if you just want to learn, then you can definitely go for it because he has explained each and everything, like even the math has been explained in there. So yeah, I'd recommend it. Also, for about the books, I would first like to recommend to you things which are non AI ml programming ones, because I found it really helpful as a programmer, we are tend to have this issue about communicating or presenting in our field, it is really important if we present ourselves properly, or our project so that we could get a job or to get somewhere around in this in this field. So there are two books actually for this or recommend. First one is how to win and Influence People by Dale Carnegie. Currently, yeah, then talk like Ted by Carmine Gallo. Those are two amazing books really help us to talk and communicate. I remember like I was in junior college. And it might also be known as school in other parts of the world. I was working as a marketing guy. So the company is not anymore right now. But it was the first step on communication and public speaking. It really helped me to communicate with people and talk with different kinds of speak different kinds of people. And in any way, anyplace. And to me like I, when I want to tell about this particular product, and a garden or let's say in a in a metro or anywhere, I would directly approach the person, talk of your chit chat and then directly Yeah, would you like to have this kind of credit? How it would be? The product was coming to courses? For sure. Yeah. So it was about that. And about the recommendation about machine learning machine learning books, I do have it right now. But it's quite big. I won't be showing it yet. So it's hands on machine learning, socket, LAN key Ross and TensorFlow, these three things are present, and it covers a lot of things, the name of that particular author is audio, or, or alien, get on
Justin Grammens 28:08
Google that and find that that into the notes. So you know, when you were talking about being so technology focused and needing to really be a good presenter, I mean, that's huge. You can't be successful only just sitting there writing code, or creating models, or, you know, it's a much more it's a very, very good career thing to be able to communicate. And then of course, of course, when you build a product, you need to understand the customer, right? Are you are you seeing, and I'm gonna phrase this but like, you know, people that probably do well in your organization, or that are working in in this in this space, they may not actually have a computer science background, like they maybe haven't gone to school for programming, because they're actually much better at, I guess the maybe the user interface. They're understanding human factors, you understand how humans communicate, you might get people that are like linguistic majors, stuff like that, that have studied language that maybe are coming in this field. I'm kind of spitballing. But would you agree with that, that that, you know, this field doesn't need to be computer scientists only this field needs a lot about other people it?
Dishant Gandhi 29:09
Yeah, definitely. Like, right now, if you are in any field, you can come to programming and you see you don't even have knowledge, you can gain it. And the plus point here is, it doesn't take any much time. Like it does not take five years to gain a particular programming language knowledge, you can just start now in a month, you are ready to try to build a program with whichever you like. And it's not just that you can even start with the front of people do start with React js right now. So in this time, it really helps each and every one like I remember I met a particular teacher, he was teaching Arabic, Hindi, English, and other languages in school. And he wanted to learn NLP also by because he thought that it would, it would really contribute in his career and also in in a way that he can learn this NLP because he knows about other languages. He can speak them, he can write them, so why not he can try to implement something with it. So he started started learning NLP and machine learning. So yeah.
Justin Grammens 30:20
And that's what I love about AI just artificial intelligence, it spans so many different fields and so many different industries, it's kind of the possibilities are endless. So it's amazing. It's amazing. It's fun, fun place to be, and well, was there anything else you wanted to cover? We talked a little bit about you and you know, your personal projects, you know, I wasn't sure about about Roscoe open source, maybe you could tell us a little bit about that community.
Dishant Gandhi 30:42
But I say the framework for building assistance on top of it. So you can use Raza for building that system. And like if you want to build any kind of assistance, you should first of all have Python knowledge and basic programming language knowledge, so that you can build an assistant on on top of it. Because this is a Python library, which you can use. And you can create different kinds of, I won't say a specific because it is just a framework where you can have you can use Bert, you can loot that classifier on top of it. And you can directly train whatever questions you have. And once we have added questions, answers, and whatever API calls or database connection you want to have, you have connected it, you can train the model, and you can chat with that particular assistant. I also have videos on it, like how you can use it how you can build forms using Raza. And that way, you can be any kind of assistance because it is really an open world where anyone can have any kind of demand like someone wants an E commerce assistant, someone wants just a voice assistant which can do lights on or off or like music home home assistant, which can open the door for you or something like that. Because I remember I worked on an auto board, automobile bot, which was for, for basic, basic stuff it would do on a particular car, it was like if you want to turn on the lights, you want to turn on the windshield, or you want to clean the window of yours. And that way it would do all the stuff like just voice command, you can just place the button or you can just say say the name of the assistant and it would do the work. Also, I created the application for laying the music giving the direction it was just an Uber, but a more wise way. So how it worked is if you want to play a music, you just say like I want to play Shape of You, it will directly it should have an integration at Spotify, or some kind of music playlist where they can access to it and play the music directly in the in that particular Android application. So if you say something like you want to get a direction, from Mumbai, to Navi Mumbai, then it will give you the directions directly opened on that portal application and the route is navigated to you.
Justin Grammens 33:03
Like you said this kind of runs, probably runs best on a on a mobile phone.
Dishant Gandhi 33:07
Yes, yes, definitely. This this particular application will run on mobile phone. But you can also integrate it with any any kind of systems, if you have that integration capability have enabled?
Justin Grammens 33:19
Sure. Yeah, I mean, on a phone, it's easy for it to just launch Google Maps, and be able to sort of plug that information in. Whereas if you're on a desktop, you're talking to your computer, those integrations aren't really there. You'd have to hand code in some ways, I guess, you know, to launch
Dishant Gandhi 33:33
that also systems and God, also in scooters, and bikes nowadays have this kind of system arrived. It shows you the direction, you don't have to have a mobile holder on top of your bike where it can show your direction. But just the indicator, it will switch back to speedometer. And there is one navigation system enable you can say it and it will drive you.
Justin Grammens 33:59
That's cool. That's cool. You've been involved in this open source. This is an open source community right at all. All the codes open?
Dishant Gandhi 34:04
Yes, yes, definitely. I share the link of this GitHub.
Justin Grammens 34:08
If anybody searches for rasa, open source, they'll find it. I found that I had not seen it before, though there must be other communities out there that are sort of doing the same thing.
Dishant Gandhi 34:18
I'm not sure about that. Because rasa is quite unique in this particular field I have seen because I, I started my career properly, like in the industry, and 2020. So when the lockdown happened, I started to learn the courses and the build portfolio and pro projects of mine by competing in virtual hackathon because at that time, there was everything was virtually happening. So hackathons were really virtualized started with virtual hackathons like MLH. Snapchat also had an iChat hackathon, where I was winning twice, and building an application and helping the community for building the application, which they had in the, in the mind, so yeah,
Justin Grammens 35:05
that's great. Giving back and doing these hackathons is another great way for somebody to get into the field. Absolutely. As they're just getting started, I have a lot of respect for people that have open source projects out there that give their time and energy to sort of help greater good and build the community. That sounds like a really cool, cool thing. I speak at a number of different conferences, whether they be technology conferences, or leadership conferences, and this would be a great topic, actually, I've written it down. And I'm like, I'm gonna take a look at this project, because it seems like anybody that if you have some Python knowledge, which is a lot of the groups that I speak to, and we've been having our monthly meetups, so it's might be a topic, or you could just sort of do a quick, you know, demo show something at one of our meetups as well. So it seems seems like it's pretty easy to sort of get involved.
Dishant Gandhi 35:46
Yes, definitely. I can showcase something in the meetup. And I could try to help everyone learn about what is rasa and how you can be done quickly with? Yeah.
Justin Grammens 35:58
Cool. Well, let's talk about that offline once for the podcast. So is there anything else you want to talk about? Before we wrap it up here?
Dishant Gandhi 36:06
Sure, I'd like to add one more thing, I really would recommend people who are in professional life, specifically in programming because we have hours and hours of coding, probably, we might also have to do something sometimes, at night, to relax ourselves, I would really recommend doing some kind of activity, which would really relax your mind. Either it'd be going in temple or doing meditation, or just roaming around around the city, birdwatching, or stargazing, even, it's, it's cool to do go for a track or dance wherever you'd like. So I myself right now, I visited birdwatching, because in Mumbai, at this particular January, and February, flamingos migrate to our city, and stay here for just two months and go back to Siberia or Iran, wherever they are, they migrate there,
Justin Grammens 37:02
that's a long way, fly. I totally agree with you know, stepping away from the code for a while. And I even think it's a great, you should be doing this. And I think you've touched on this to start your day off with this. One of the big epiphanies that I've had over the past couple of years is is I used to just open my laptop, open up, you know, email, and then just start my day that way. And I did a complete mind change in 2019 were I really focused on exercise first of all in the morning, because it just felt a lot better after I got through the exercise routine, did a little bit of studying around some topics for me was artificial intelligence and machine learning. So learn a little bit, do some journaling, and some meditation, and then look at what my day is going to be like, plan out the three or four things that I want to make sure I get done, and then open the laptop and start so I've been way more productive and way more happy. Because of that, just having a good Good morning routine, but exercise and doing something else is a huge part of it. For sure. Well, good, good. Well, I appreciate the time again today. Sounds like you guys are doing some really fun things at Horry and you've been doing some cool stuff in the open source community and maybe we'll have you at one of our meetups in the future but Deshawn Connie, thank you so much for being on conversations with applied AI or to keeping in touch with you.
Dishant Gandhi 38:14
Thank you so much.
AI Announcer 38:17
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