Return to site

21. George Vettath on complex big apps and the rise of low code products

· podcast,AI


broken image


Podcast with George John Vettath Part 2


Podcast with George John Vettath Part 2


George shared his views on how low code can enable greater agility for iterating and building complex applications. Gartner predicted that by 2024, 65% of apps will be low code. George predicts that all apps in the next decade will become low code. Business leaders will find that the first specifications that come from who is conceptualizing the application are only 80% ready. When companies are trying to complete their project specifications from 80% to 100%, They may have to go through maybe up to 40 iterations before they can get to a point where say they are happy with the application. Now a simple app will require maybe 10,000 or 20,000 lines of code. Yes. A complex app will require 300,000 lines of code. 30 times 50 times the size. So Fortune 500 companies are heavily adopting enterprise-class, low-code platforms and these low-code brands like Mendix and Outsystems emerge.

Notes and transcript can be found here:

[00:00:00] Andrew Liew Weida: Hi, everyone. Welcome to the AI of mankind show where I share anything interesting about mankind. I'm your host for this season. My name is Andrew Liew. I work across four Continents and 12 international cities. Also, I work in tech startups across a range of roles from selling products, making customers happy, figuring out fundraising, making finance tick, building teams, and developing sticky products. Apart from building startups. I've also worked in fortune 500 companies as a chief data scientist or technologist or people leader. You can call me Jack of all trades or master of learning. I hope to make this podcast show a great learning experience for us In each season, there is a series of interesting things that I invite guests to share their views about their life and interests.

[00:01:09] Andrew Liew Weida: Now let the show begin.

[00:01:26] Andrew Liew Weida: In the previous episode, George talks about his backstory. Along the way, he shared his views on how covid19 has accelerated digital transformation. This episode continues the part 2 conversation with George and George shared his views on how low code can enable greater agility for iterating and building complex applications. Let's continue.

[00:01:44] Andrew Liew Weida: You mentioned there were four things that is happening at the same time is that we gotta deal with people, whether is it the user, the developer of whoever is involved in the project. We gotta think about return of investment. We gotta think [00:02:00] about the technology architecture. We gotta think about compliance. And so coming back to you, so you mentioned about that's why low code can enable all these four different problems to be solved at the same time. Tell us how you know

[00:02:14] George Vettath: So in short, again, just to remind low-code is getting things done. Let's say at one third or one fifth or 1:10 the effort. So that's anything that you can do faster, better is going to be useful. Now, how will it help? Let's say, for example, if you want to check out your, how your business strategy is going to get affected with a new plan that you're going to do, I will, anytime I do a new initiative, I would not try to write a code, to inspect document and try to build. They have the ideal way is to take a low code utility, build something quickly mock it up show it to 4 to five people who are going to be stakeholders, verify whether the intended benefit is going to happen. So that's how low code is helping in business strategy.

[00:02:51] George Vettath: Same happens with change. For example, when you're trying to change you, you should be waiting till the end of the application. Before you try to change end users while the development is going on, you should be showing semi mocked up applications to end users and gearing up for the change. And here again, low code helps because the way low code does. You auto develop 80% applications to the users. It'll seem like it's automatically developed. It's already ready. Reality is only some few and UIs. So the change management can also be done in parallel same with compliance. Now, compliance is a change actually is simple software, but fairly I call it a lot of work. So you wanna build a lot of screens. You wanna build a lot of reports. Again, it's not a manual work from a coding perspective. And here also you can deliver more than less now. As far as architecture is concerned. Now, every time you don't want to rework the architecture for every application. You don't want an architect to come in say, okay, this is going to be the UI layer, middle layer, the backend, et cetera, and then rebuild and restructure. That's why so many applicants have submitted the platform. Low code platform comes with its architecture. It's scalable, intrinsics cloud ready, intrinsic. It has security intrinsic, and they're getting better and better. So by going into a low code platform, you just focusing on the problem at getting most of it done and completing only the bits of the low code platform cannot do.

[00:04:09] George Vettath: So low code platforms are changing things so much so that Gartner is predicted that by 2024, 65% of apps will be low code. My personal prediction is all apps in the next decade will become low code. And that's why for example, stragility is so important for us. We are using stragility for everything that we do.

[00:04:26] George Vettath: I believe that over time more and more complex application gets up and there's definitely a need to use low code. Now, the interesting part you mentioned about rapid prototyping, maybe you can help to educate the audience out there because some of them, they are digital designer or the user interface, user experience designer, where they use those kind of like Figma or sketch Adobe where they pen and paper wire frame, it draw the buttons and the circles and the gesture to do rapid prototyping. How is that different from like you mentioned low code production.

[00:05:00] George Vettath: I have done prototyping for a number of products and I found that the biggest problem. As much as it is UI is very crucial, very important. I'm not saying it's is moving the completeness of the specifications from 80% to 100%.. That's a very difficult part because you may draw around nice wire frames. You may paint the screens. You may write up, look, this is what's going to happen in this form. This is how the landing page will be. You may get all that and get a nice UI. But you find that first specifications that come from who is conceptualizing the application is only 80% ready. And from 80 to 100, you may have to go through maybe 10, 20, 30, 40 iteration before you can get into a point where say you are happy with the application. Handling these iterations are difficult because you gotta get show the prototype, give, get feedback, put that back into the product, show it, write the rules, go to iterations. And for that, if you use a painted one, the problem is every time the UI design has to keep changing. So what I would do is rather let's leave the UI a little down the line, let us first iterate and prototype how the application would look from a navigation perspective, from a process perspective, the screen perspective, from the content perspective, from the role perspective, then look at it as what should the parts where different roles are going to use the application. So if I have end user doesn't have software, I would then take care, make it as user friendly as them. And if I take it for a someone is familiar with software. Maybe I'll just make it a little simpler. But for the application administrator, I'd give a regular screen automatically generated screen form, and you'll be surprised 80 to 90% of an application use a regular font. It's only 20% of the applications that's a surface where is being seen by the end users. That needs to be beautifully painted.

[00:06:45] Andrew Liew Weida: So you're saying that like typically those people who are the UI guys. They do wireframe. They do paper form is pretty much 80% of what the requirements are. But when you put into production means you, you put the code, you put the business logic, you compile it and you host it on the cloud and then when you finally test it, maybe it's different. That 20% is so crucial that most people were to do it waterfall instead of agile on those traditional like get a few developers and code it. That 20% is critical because when you say that, when you wanna change, it's so painful. So hard.

[00:07:19] George Vettath: Yeah it is so painful, so hard. So it wasn't possible. So the approach you kept you take the right one when you do the UI first at the time when low code was not there. When low code was not there, I would get the UI done first, but with low code there, I can iterate daily. I can iterate 40 times within 40 days. They do only one iterate. Do keep iterating until you get it right, then do the UI and then complete, so that's a different approach in low code, which I'm recommending now. And only those who are familiar with low code will understand or appreciate this because I can literally keep improving the application on a daily basis and do the UI automatically using low codes.

[00:07:56] Andrew Liew Weida: Let's look at the other side of the argument or the other side of like, there's a bunch of all these software engineers, and they say that now we have agile. We can just code on the fly. We don't really need low code. What do you have to say about that?

[00:08:07] George Vettath: Agile is needed for coding, absolute essential, fully and agile properly. We use agile too. It's just that it has be done in the past where coding is to be done. Okay. Now, Let's say, for example, I'm not talking about simple apps. I'm talking about complex apps here because that's my experience.

[00:08:23] Andrew Liew Weida: Help the audience to understand what's a simple app and what's a complex app. Just to give us a very simple, let's say using HR like leave or claim. Just give an example.

[00:08:31] George Vettath: Let's say for example, you just have a simple app, which does just recognize this. It does leave application for you. Mobile leave application. One screen, you create a leave application. Maybe you can just view your application. But a complex application is let's say a recruitment process end to end whereas a request, then there is a candidates submit and then you go through a whole process of eight or 10 steps because someone is onboarded. That's a complex application.

[00:08:56] George Vettath: Now a simple app will, will require maybe 10,000 or 20,000 lines of code. Yes. A complex app will require 300,000 lines of code. 30 times 50 times the size. Now, when you're doing that, suddenly you're dealing with a lot of complexity. Okay. And the problem is the companies are littered with small apps. SaaS was popular recently, but it's going to go back to connected apps, larger apps. And that's when again, low code is coming to becoming important. Now, therefore, when you want to build a 300,000 line code application, you have a development app.

[00:09:29] George Vettath: Now, if you don't have low code, you will be putting a 50 member team. If you have low code, you could do with three member. In both cases you need agile. Yes. Agile is still a method by which you plan your work, then you buffer it and then you introduce it one by one in pieces to to the team and the team takes 15 day sprints and complete it. That's still lead it. It's just still. That the quantum of work is lesser with low code because out of the 300,000 lines of code, maybe 260,000 of them would've been automatically generated in low code. Whereas in conventional you're writing almost say 200,000 lines so therefore agile is needed in both situations. It's just that the quantum of agile can be done, three sprints or two sprints instead of 20 sprints.

[00:10:12] Andrew Liew Weida: So you're saying that a complex, recruitment apps where you have multiple approval, multiple sequences, and 300,000 lines of code and you need 50 software engineer. Now with low code, you still generate, let's say 300,000 lines of code, but you can do it faster, better with only five engineers. So if it's such a fascinating thing, then why hasn't Csuite executives or corporate leaders begin to try using low code as an approach?

[00:10:41] George Vettath: So the answer is essentially in smaller apps, a lot of low code is there. So you have even Google, Microsoft, everyone's got something on low code. So you've got app sheet, Google, you've got power apps or Microsoft. So lots of apps, small apps getting out. Now in the big apps, the complex like apps, low code platforms are few in number. Okay. So the names of the companies are names like Mendix, outsystems. We like to put ourselves in that category now, such apps are expensive. So Fortune 500 companies, companies in the USA are adopting enterprise class, low code platform, very heavily. Take, for example, the where the valuation of these five companies include service now, for example, just look at the trajectory of the valuation of these five companies. They're not talking about million dollar valuation, they're talking billion dollar valuations all shot up in the last few years. That indicates how much of adoption is happening among enterprise class, low code platforms.

[00:11:40] Andrew Liew Weida: Now they are not concerned, They are excited and they are giving a try at these low code enterprise complex solutioning systems. As you've also mentioned, they are expensive. Give us a range of numbers so that we know what does expensive means. Are we talking about a million dollar, a $10 million, a hundred million dollar.

[00:11:58] George Vettath: I'd like to make it simpler. Now I can give an example from right in Singapore, we had a fund management company and they had got six quotes from six different vendors, all quoting between say $200,000 and $1 million. When we met the CEO, we said, we promised we do it half the price of the lowest vendor and half the speed of the fastest vendor. And we did that, sign the contract and delivered what was expected to take eight months, we deliver it in two months. So there are a lot of examples like this, where all the low code vendors will show real life examples where they've reduced costs significantly and timeframes significantly. So adoption rates are picking up quite rapidly.

[00:12:36] Andrew Liew Weida: You're saying that relative to traditional coding method or systems solutioning method, it is still cheaper . So when you mentioned George yours is even cheap more cost effective than the existing big brands, low code that we're talking about. The Mindex, the ServiceNow.

[00:12:52] George Vettath: That's right. So what we do is, so they cater to the fortune 1000 companies, we cater to the companies who don't have so big budgets, but the need is there for everyone.

[00:13:01] Andrew Liew Weida: Hi everyone, thanks for tuning into this episode. We have come to the end of part 2 with George . In the next episode, we will continue with George on part 3 which he shared his advice on how business leaders gonna build a business case for using low code in digital transformation to the board . On top of that, he offer his view on which industry handles the most number of complex applications. Finally he will share with us his view on how to enable AI with low code .

[00:13:25] Andrew Liew Weida: Hi guys. Thanks for listening to this podcast. If this is the first time you are tuning in. Remember to subscribe to this show. If you have subscribed to this show and love this. Please share it with your friends, family, and acquaintances. See you later and see you soon. .