Building Self-Sustaining Systems in Tech Organizations

00:10 - 00:15
My name is Matt, and I will be talking to Khaled Souf about industry dynamics and continuous learning and adaptation.

00:18 - 00:20
Hey, Khaled. How are you doing?

00:20 - 00:22
I'm doing good, Matt. Thank you.

00:22 - 00:23
Thank you for having me.

00:23 - 00:25
Yeah. Happy happy, happy to have you here.

00:25 - 00:27
I mean, first time in, Geneva.

00:27 - 00:33
We just recently talked that the vibe is interesting, but it's more of a French city. Right?

00:33 - 00:38
Yes. It's actually like, well, the the the Geneva people, they will not like what I say, but

00:38 - 00:44
usually, we we say that Geneva is part of France, not of Switzerland because of the way they

00:44 - 00:47
work and for many other reasons.

00:48 - 00:50
So Switzerland start after Geneva.

00:53 - 00:57
Haven't I haven't known that, so that's a good to good, good thing.

00:58 - 01:01
And we are here to discuss the the tech stuff in general.

01:01 - 01:06
We already had some discussion, and I have to stop you because you have so many interesting stories, to be honest.

01:09 - 01:14
And, like, what caught my attention, when I look at your profile and we discussed the potential

01:14 - 01:19
podcast before, it's like you are, I would say, around 20 years in tech, right, already?

01:20 - 01:24
Well, not, like, more than 15 years. More than 15 years. Okay.

01:25 - 01:29
And you have, because you work as a consultant, so you have a lot of experiences.

01:29 - 01:34
So if I when I check the LinkedIn, there's, like, a number of 36 experiences, which is, like,

01:34 - 01:38
more or less different positions, different companies that you consulted.

01:38 - 01:42
So we have a lot of stories that I would love to hear today.

01:42 - 01:44
Yeah. We I'm I'm what we call the nomad.

01:44 - 01:46
So so it's like I'm always switching.

01:46 - 01:50
And, with with these years, it's become even more hectic because sometimes I can have, like,

01:50 - 01:54
2 to 3 customers in a week. So Okay.

01:54 - 02:00
So it's more and more and, like, today, the, the mandate, they are smaller and smaller.

02:00 - 02:02
So it's like I can stay, like, for a few weeks.

02:02 - 02:05
Sometimes it's for months for for for customer.

02:05 - 02:09
But most of the case, less than a year. So I'm always switching.

02:10 - 02:15
So nice. So you have, been there, done that, have seen a lot of things probably in your life.

02:16 - 02:25
And 1 thing that, stuck in my mind when we, talk, you work for 1 company in, in Canada, and

02:25 - 02:31
we were discussing scaling teams, scaling products, and you told me, like, really interesting

02:31 - 02:37
story and your lessons are based on that on scaling the the product or scaling the technology behind the product.

02:37 - 02:39
So maybe could you elaborate on that?

02:39 - 02:47
Yes. So what happens is actually, I joined that team as, well, we tailor it actually a mandate for myself.

02:47 - 02:51
So it's like I've been discussing with 1 of the manager about the issues that have.

02:51 - 02:57
And and I told him, okay, this subject are actually mainly I can I can help you with that?

02:57 - 03:04
So we started with this subject. 1 of them is actually migrating the, let's say, the the main

03:04 - 03:09
ecommerce applications to a new cloud provider for different reasons.

03:10 - 03:17
And, in the middle of that migration, well, they they they told me that actually Black Friday

03:17 - 03:21
and Cyber Monday is coming. So let's stop that.

03:21 - 03:29
Let's, let's, pass that, let's say, big wave of customers coming to to the website and then

03:30 - 03:32
buying a lot of stuff, and then we can go there.

03:34 - 03:39
So what I discovered with the with the with the with the with the ops that that were there,

03:39 - 03:46
that we've been when we've been working, well, we check the traffic over the few years back.

03:46 - 03:50
We discover that the traffic in these days is like 500%.

03:51 - 03:54
Than, let's say, a classical day.

03:55 - 03:57
So we needed to prepare that.

03:57 - 04:07
We checked also whether the application that's gonna be, solicited and mainly working more than the other.

04:07 - 04:12
So we can know how which 1 should be scaled and which 1 they shouldn't be.

04:13 - 04:19
And, well, we've prepared we've prepared the work because you already have, infrastructure as code.

04:19 - 04:25
So it's like changing the infrastructure is easy and, you know, you can we can keep it as it is.

04:25 - 04:28
So what we did is, like, we said, okay.

04:29 - 04:36
If you need to scale, you can't use the, what you call, the automatic scale stuff that you can find in cloud. Why?

04:36 - 04:38
Because it's already too late.

04:39 - 04:47
The the time that need to to prepare virtual machines or maybe virtuals, let let's say deploy

04:47 - 04:51
application, we're gonna we're gonna lose something if we do that.

04:51 - 04:53
So what we did is actually said, okay.

04:53 - 04:57
We know for sure at midnight, they go we're gonna have that.

04:57 - 05:02
So let's prepare the day before. We're gonna, prepare instances.

05:03 - 05:06
We're gonna prepare, like, already the application.

05:06 - 05:11
And then after the cyber Monday, we're gonna we're gonna shut them down because it's, it's,

05:12 - 05:15
let's say, it's very specific.

05:15 - 05:19
I will not, like, buy a virtual machine for 3 years or for 6 months.

05:19 - 05:26
I know for sure that I have, like, 2 days that so so I need to scale only for these 2 days.

05:26 - 05:32
Let's say even I included also the weekend because, Souf know, we have Black Friday.

05:32 - 05:37
You have then Saturday, Sunday, and then the Cyber Monday.

05:37 - 05:40
So what we said, okay, let's scale it for 4 days.

05:41 - 05:44
We're gonna stay on-site for that weekend.

05:44 - 05:46
They're gonna pay us for that for sure.

05:49 - 05:56
And, and, well, we did it, and it works just fine because we already have information how much

05:56 - 05:57
is gonna be the traffic.

05:58 - 05:59
And that's that's cool also.

05:59 - 06:07
We, we've made a lot of alerts Souf so we can know if if if it attains a certain level, we can

06:07 - 06:09
then maybe add more machines in that.

06:09 - 06:11
So that's why we've been we've been there.

06:11 - 06:14
So everybody was checking, and it was really nice.

06:14 - 06:15
This is 1 type of scaling.

06:16 - 06:22
This is the scaling that you have, like, multiple users that is coming up, and you try to scale

06:22 - 06:24
for for an event or something.

06:24 - 06:30
It happens a lot, especially when when people are buying for, for holidays, for Christmas.

06:30 - 06:34
This is happens a lot for, like, Black Friday and Cyber Monday.

06:34 - 06:37
Usually, it's like what they buy is actually for for Christmas.

06:37 - 06:42
And, and then, there is other type of scaling.

06:42 - 06:52
So the other type of scaling is, let's say, okay, we have market and we try to scale the product to multiple countries.

06:53 - 06:55
This is a different type of scaling.

06:56 - 06:59
This 1 needs a different, different things.

06:59 - 07:06
So 1 of them is actually most of the time you're gonna need to readapt your application, extract

07:06 - 07:11
the 1 that actually common, the part that are actually common and try to figure it out this

07:11 - 07:13
specific 1 for the country.

07:14 - 07:20
And then Souf create some kind of contracts, let's say, of these specific parts.

07:21 - 07:24
And every time you have a new country, you need to implement it.

07:24 - 07:26
This is in in the code side.

07:26 - 07:31
When we go to the deployment side, what happens is actually we say, okay.

07:32 - 07:39
Some people, they're gonna tell you, okay, let's do 1 application for all multi tenant, multi, countries.

07:39 - 07:42
Well, usually, I don't agree with that. Why?

07:42 - 07:47
Because it doesn't make sense that somebody in France need to use the application.

07:48 - 07:51
He gonna call the website that actually in the United States.

07:51 - 08:00
And especially, and also it's, it's it's a risk because if the the application is shut down

08:00 - 08:05
for for in USA, everything is shut shut down in every country.

08:06 - 08:13
So in my opinion, the best way is like taking that application, deploying a new instance not

08:13 - 08:18
far away and even better if it is deployed in the country itself, if there is a region for that,

08:18 - 08:21
if there is like, like a place in the cloud provider.

08:21 - 08:24
So it can be not far away from customers.

08:25 - 08:30
We can manage the traffic because we know how much we're gonna have traffic for that country.

08:31 - 08:34
And in the same time, it's completely independent.

08:34 - 08:38
So it's like when we deploy a version, it depends on the country.

08:39 - 08:44
The, the drawback for this type of solution is the common part.

08:44 - 08:52
So when you try to upgrade or update or maybe add new features for the common part, you need

08:52 - 08:54
to deploy them country by country.

08:55 - 09:03
So 1 of the things that I prefer in this side is, like, using feature flapping, you know, feature toggling.

09:04 - 09:10
So we can, like, we for the common part, we can, like, activate, deactivate certain feature

09:10 - 09:17
and then activate them in the other country if the country is already like, let's say, we already

09:17 - 09:20
prepared everything for that, and we do it slowly, slowly.

09:20 - 09:22
You know, I'm not a big fan of big bang.

09:23 - 09:29
I always try to do things slowly, because in my opinion, something that is already working in

09:29 - 09:41
production need to be evolved, not we we don't need to redo it. It doesn't make sense. It's already making money. So

09:42 - 09:46
Makes sense. So I assume you don't deploy on on Friday very often.

09:47 - 09:50
Well, yes. I prefer deploying on Monday.

09:51 - 09:58
Or I prefer deploying when there is less traffic and we have people for the next 2 to 3 days

09:58 - 09:59
so we can manage the issue.

09:59 - 10:02
So it's like, Wednesday, it's a good day.

10:03 - 10:08
Thursday Friday, they are not, like, the best 1 to choose. Monday, the best 1. You know?

10:08 - 10:13
It start the journey with, like, with, like, a big big thing like that.

10:13 - 10:19
And, and depending on the context on the, on the on the traffic, like like I told you, sometimes

10:19 - 10:25
when you work in ecommerce, usually when we deploy, you try to deploy, you check how much is

10:25 - 10:32
used the application, and then you try to figure it out like, like, the the the the time that

10:32 - 10:34
nobody is using it or maybe less people.

10:35 - 10:38
You communicate on that, you know, what you call a change management.

10:39 - 10:43
Like Souf communicate about the new feature that is coming up and then you say that we will

10:43 - 10:47
not be available from that hour to that hour.

10:47 - 10:54
And you know for that hour for for these, like, area of hours, nobody gonna use or maybe little

10:54 - 10:58
of people, like 2 or 3 or maybe less or maybe even nobody.

10:58 - 11:01
So there you can manage the risk.

11:01 - 11:02
And this is the thing.

11:02 - 11:07
So if there is a risk, there is like multiple ways to to do it.

11:07 - 11:08
Solve it, if you can.

11:08 - 11:10
If you can solve it, you try to manage it.

11:11 - 11:15
And this is 1 way of managing the risk of actually deploying.

11:16 - 11:20
And and talking about the risk, we discussed really interesting, things.

11:20 - 11:23
To be honest, I haven't heard it before from my guests in the podcast.

11:24 - 11:26
You mentioned the term of chaos engineering.

11:26 - 11:28
Oh, yes. That's, that's a good thing.

11:28 - 11:31
So, this is actually a thing that I discovered.

11:31 - 11:37
So so it's like, when Netflix was like the only platform that is existing and everybody was

11:37 - 11:45
talking about, they have they have like very interesting way, to deploy in production. 1 of

11:45 - 11:50
the things they were using microservices, but now everybody knows that microservices, there

11:50 - 11:53
is like, trade offs on that.

11:53 - 11:54
We can go back to that.

11:55 - 12:01
But, the idea that I like it what they do is, like, there is a story about the whole region

12:01 - 12:07
that shut down, of Amazon in United States, but Netflix still working.

12:08 - 12:11
And mainly, at that time, Netflix was only on Amazon.

12:12 - 12:14
Right now, they are poly cloud. That's different.

12:15 - 12:18
So everybody was looking how do they do it.

12:19 - 12:25
What they actually did is actually they have something that they call, like tools that they

12:25 - 12:30
call cow cows monkey or the 12 monkeys for the list of all the tools.

12:30 - 12:36
And what these tools do is like they go to production and they shut down services.

12:37 - 12:44
And then from there, the engineers, the ops, the devs, they're gonna try to figure it out how

12:44 - 12:51
the system behave in a degraded mode and try to manage that risk like like usual.

12:51 - 12:55
So when you can't avoid the risk, you try to manage it.

12:56 - 13:01
And, 1 of the tools that comes from that is, like, what Souf call circuit breaker, for example.

13:01 - 13:08
It's like when you have an exception or an error in in in microservice, you try not to propagate it to other service.

13:09 - 13:12
And you try to have, like, some kind of cache.

13:12 - 13:19
So when somebody is calling, he get the last information, even the when when the the service is not available.

13:19 - 13:23
This is 1 of the solution they figure it out by using, chaos engineering.

13:23 - 13:31
So the idea of chaos engineering is, like, let's shut down, like, random service and see how

13:31 - 13:40
the system is behaving in so we can make it behaving in a way that keeps working even if it

13:40 - 13:42
is in in the degraded mode.

13:42 - 13:48
And you mentioned the microservices, so microservices versus Monolith and because, like, the

13:48 - 13:53
spartigmal, it's changing over the last over the last years, everybody was talking about the

13:53 - 13:59
microservices, And I think like this or last year, the people are starting changing the approach.

13:59 - 14:01
So what is your what are your thoughts here?

14:01 - 14:07
Okay. So, the thing about microservices is like, we started with the hype, so everybody's going there.

14:07 - 14:15
Everybody but the thing is when Souf are a CTO or an architect or somebody who have a technical

14:15 - 14:21
vision, we understand and we know for sure that everything is coming is coming with 2 things.

14:22 - 14:24
It's coming with, with benefits.

14:24 - 14:26
And there is always drawbacks.

14:26 - 14:30
And at that time, people were were focusing on benefits.

14:30 - 14:32
Like, let's split it in small modules.

14:32 - 14:35
Let's split it in a small service, and then it's easy to deploy.

14:35 - 14:37
It's easy to to undeploy.

14:38 - 14:39
And the risk is actually mitigated.

14:40 - 14:42
But the truth is actually is not that.

14:43 - 14:51
First thing is, like, it's simpler for developers to do it, but it brings multiple switch context.

14:52 - 15:00
It brings also a lot of complexity on the upside because you need to deploy like 100 services, microservices or something.

15:01 - 15:06
So you need to create some kind of strategy so you can provision that as fast as possible.

15:06 - 15:11
And also with all the resources, you need to industrialize that.

15:11 - 15:13
So there is a lot of work to do.

15:13 - 15:15
So this is 1 of the thing.

15:15 - 15:22
The other thing also is, like, when you try to trace back something that is coming from the customer, it's very hard.

15:22 - 15:30
So you need to centralize that in some place so you can trace back all the, let's say, all the

15:30 - 15:35
what happens when there is an issue Because it's not easy to figure it out the issue when you

15:35 - 15:37
have multiple, places that you need to check.

15:38 - 15:42
So you need some place to to do that. This is 1 thing.

15:42 - 15:50
So now the other and the biggest problem in my opinion is, like, people tend to go to microservice

15:51 - 15:54
very early when they don't understand their business.

15:56 - 15:59
What I what I recommend usually is, like, we go to Umorate.

16:00 - 16:02
We stabilize the business side.

16:02 - 16:10
Usually, like, I'm I'm I'm I'm a big fan and also a big I have a knowledge on domain driven designs.

16:10 - 16:14
What what I what I go is like monolith and then you go to modular monolith.

16:14 - 16:16
So you can split it in small models.

16:16 - 16:23
And then if there is a reason to scale, especially if you need to scale that part, then you

16:23 - 16:25
extract it as a microservice.

16:25 - 16:30
So what I prefer is, like, going to something hybrid between monolith and microservice.

16:31 - 16:35
But you need always to check if the business need it.

16:35 - 16:37
We always go to the same stuff.

16:38 - 16:43
People tend to go to the technical solution without looking to the business need.

16:43 - 16:45
Always looking to the business need.

16:45 - 16:47
And this is this is actually the role of CTO.

16:48 - 16:51
He need to understand exactly what the business need.

16:51 - 16:57
He need to be with the business so he can translate that in a technical vision.

16:58 - 17:03
If he can do that and he goes with the hype, it doesn't make sense actually.

17:04 - 17:10
They they will be like like a big wall between the business and the the the technical side.

17:10 - 17:12
Let let's let's be frank about that.

17:12 - 17:16
Without the business, we don't have jobs as the techies.

17:16 - 17:22
So we need to understand that everything that we learn, tools, techniques, anything that we

17:22 - 17:31
learn in tech is actually something that can be helpful to the business people so they can bring value to the world.

17:31 - 17:32
This is how I see it.

17:32 - 17:39
And this is the point of that the CTO need to be the pillar for that for the team.

17:40 - 17:44
I I think what you are saying, I I absolutely agree with that 110%.

17:45 - 17:49
At the same time, I have a feeling that to become, like, a good leader, good CEO, and listen

17:49 - 17:53
to the business, it's the same as Souf I don't know.

17:53 - 17:58
Have you played any instrument in, in the past? Well, no. Okay.

17:58 - 18:05
But the example is, like, usually when people start to playing guitar, they start to do, like,

18:05 - 18:10
a hard solos, like, really difficult stuff, like, adding a lot of effects, a lot of different guitars.

18:11 - 18:16
But when they are experienced, they use 1 guitar, they play just a few notes, then then don't

18:16 - 18:19
use a lot of stuff because they play what is needed for the Souf.

18:20 - 18:22
Right, to be, like, value added.

18:22 - 18:24
So the same is of the with the role of the CTO.

18:24 - 18:29
Like, you need to check different stuff and then come back to the fundaments. Right?

18:30 - 18:37
The the best song that you hear usually is, like, 1 voice and maybe 1 instrument like, like,

18:37 - 18:39
an acoustic guitar or something.

18:39 - 18:41
These are the best sound that that you hear.

18:41 - 18:47
And usually people, they tend to to go to there than multiple instrument with a lot of stuff.

18:48 - 18:57
It's exactly. Exactly. So, so the the next question that I want to ask, I'm just always wondering

18:57 - 19:02
after so many years and experiences that you have got, what was the toughest experience that

19:02 - 19:04
you have got and how this shaped you?

19:04 - 19:06
Well, there there is a lot of them.

19:06 - 19:14
So I can because I'm a consultant, so I like I have a lot of tough, let's say, experiences.

19:15 - 19:17
But, 1 of them is actually

19:20 - 19:23
I remember that I worked with a team with 2 seniors.

19:23 - 19:25
I was a junior that time.

19:25 - 19:32
And this is actually 1 of the first experience that I've seen, and they were actually perfect jokes.

19:32 - 19:34
They were like rock star or something.

19:34 - 19:42
And that actually made me realize that I, I told myself when I get to that level experience,

19:43 - 19:44
I will not be like them.

19:45 - 19:50
And and actually, I hope that's the case. No. You never know.

19:50 - 19:51
Somebody is actually rude to somebody.

19:51 - 19:53
It can be jerk for some people.

19:53 - 19:55
So but I'm trying to not to go there.

19:56 - 20:01
But, what I'm trying to say is, like, I was at that time, I was junior.

20:01 - 20:04
So when I asked for help, I never get it.

20:04 - 20:10
But now I'm actually, when somebody needs help, even if something like basic and sometimes some

20:10 - 20:13
people they're going to, they're going to tell you, hey, go to Google or something.

20:14 - 20:19
I will tend to help them first and then tell them that before going back to me, these are the

20:19 - 20:21
steps they need to go.

20:21 - 20:23
Like, they need to do that, that, that, that.

20:23 - 20:30
Then when they arrive in that step and they don't find a solution, then they can get back to

20:30 - 20:33
me so we can make the time more efficient for both of us.

20:34 - 20:37
And it's also it's a learning experience for them.

20:37 - 20:44
So it's like it's easier, Souf know, when people learn something by themselves, they tend to not forget it.

20:44 - 20:46
And that's actually a way of work.

20:47 - 20:50
And this is the difference between knowledge and wisdom.

20:50 - 20:55
And this is something very, very interesting in tech industry right now.

20:55 - 21:02
There is, there is a lot of people with 1 year multiplied by 10.

21:02 - 21:05
But there is a few people with 10 years of experience.

21:06 - 21:11
So I'm trying to say is like the 1 that 1 year they are repeating the same year.

21:11 - 21:19
They are not like trying to learn something new to, to improve themselves and to go beyond that.

21:20 - 21:25
For those who have 10 years of experience, usually they are people to tend to to know.

21:25 - 21:30
What happens is actually the 1 that have 1 year experience, they develop what they call knowledge.

21:31 - 21:37
So they know how, how it works for that language, for that specific context.

21:37 - 21:41
But if they go outside of the context, they're gonna switch back.

21:41 - 21:49
But for those who have 10 years of experience, they develop mechanism to learn new things.

21:49 - 21:53
And especially, you know, that in the tech industry, it's moving too fast.

21:53 - 22:02
So you need to develop some kind of ways so you can learn fast things and you can understand

22:02 - 22:08
fast faster than most of the people these things, especially when you get old. You know?

22:10 - 22:18
So, so the thing is, there there is, like, you know, going to conference, having having a network

22:18 - 22:23
of people that you can ask for for for the best resources that you can start with.

22:24 - 22:30
Also reading yourself books, on the tech side and also outside of the tech side, especially

22:30 - 22:35
if you are working with business, trying to understand how the other people, working, behaving,

22:36 - 22:41
especially when you take a big role, a role of leader, especially in the tech side, you need

22:41 - 22:47
to understand how the business thing, how the marketing thing, how the sales actually thing,

22:47 - 22:53
and try to figure it out how you can make everything is actually, working for them because they

22:53 - 22:55
are your customers at the by the end of the day.

22:56 - 22:59
And also they give you feedback, and this is actually the most important part.

23:00 - 23:05
Because, like like like anything, you need feedback.

23:05 - 23:10
You need to create that feedback loop which actually can give you so you can understand your customers.

23:10 - 23:15
And from from there, you try to build something, between between you and the business.

23:15 - 23:16
I'm not talking about process.

23:16 - 23:19
I'm talking even in the business, vision.

23:19 - 23:22
In my opinion, the city who can give his feedback or say, okay.

23:22 - 23:23
Let's go to that way.

23:23 - 23:25
Let's try that because that's easier.

23:25 - 23:28
And then we check how the customer behave for that.

23:28 - 23:35
You know, it's like, when you are dealing with people as customers, we need to understand that

23:35 - 23:39
it can go different than what you think in your head. You know?

23:39 - 23:43
Like, you can give a product that you say it will not work and then at some point, you understand

23:43 - 23:50
that, that that that product actually worked well for for custom. 1 of the story that I can

23:50 - 23:52
give you I can give you is like about Slack, actually.

23:53 - 23:58
Slack, when they started, they were, they were a gaming company. They they created games.

23:59 - 24:05
What happens is actually in these games, they created that channel so that people can talk between each other.

24:05 - 24:10
And now the game is completely gone, and they switched their business to go completely on that

24:10 - 24:12
side, and it's working very well.

24:12 - 24:16
So this is 1 of the story that somebody who is actually listen to the customer.

24:17 - 24:22
He understand that the main, business area, it's not actually the game itself.

24:22 - 24:23
It doesn't work for the games.

24:23 - 24:26
It works actually for the other other side.

24:26 - 24:34
Like, few days back, I was discussing also with some person, I will not disclose the the the

24:34 - 24:37
thing, but we were discussing they are moving to different Souf.

24:38 - 24:40
And they are using AI for something.

24:40 - 24:48
And they discovered that, and they tried using for for for another, let's say, another way.

24:48 - 24:55
And they discovered that the other way is actually more interesting to the customer than the first 1.

24:55 - 24:56
So they completely ditched it.

24:56 - 25:02
And this is actually 1 of the biggest advice that I can give to, to start up. Listen to your customer.

25:03 - 25:09
See where where actually you bring value and see where they think where where the customer see the value.

25:10 - 25:16
Sometimes if you are too focusing on, okay, this is the value in your head and you don't listen

25:16 - 25:18
to the customer, well, it doesn't make sense.

25:18 - 25:23
If somebody doesn't want to buy your product, you can't force them to do.

25:23 - 25:27
But you can change your product to the needs that that he have.

25:27 - 25:34
So you need to listen to him, about the and get as much as possible feedbacks about that.

25:34 - 25:35
That's why there is a lot of surveys.

25:35 - 25:42
There is a lot of things that that usually, companies do so they can get the feedback as much as possible.

25:43 - 25:48
Yeah. I think being client centric, like, especially in this age when there's, like, a huge competition.

25:48 - 25:52
Like, there is so many solutions, like, coming live.

25:52 - 25:58
I heard 1 story what changed my mind recently. 2 years ago or last year, I was in Tel Aviv with

25:58 - 26:03
my cofounder and, you know, we don't do those, podcasts as as as we are doing now.

26:03 - 26:08
And he was talking with 1 guy, and I don't know if those those podcasts are maybe outside, but

26:08 - 26:09
the guy sold a company.

26:09 - 26:13
He's, like, filthy rich, filthy rich, and he's starting another company.

26:13 - 26:16
And he's telling about the process that what he is doing.

26:17 - 26:21
And, you know, he he said, like, his goal is to talk with, I don't know, 70 people.

26:21 - 26:25
He has, like, a customer interview. Yeah.

26:25 - 26:26
Like, he's doing customer development.

26:26 - 26:31
He has a set of questions, and the first thing that he's doing, he said, like, I need to talk

26:31 - 26:34
with the potential clients, and he's doing his on on his own. Right?

26:35 - 26:40
So, like, you need to go to those basics, and, like, I got really expecting inspired by him.

26:40 - 26:46
So I started to talk a lot of the people, like, to save those data and then to discuss with my team.

26:46 - 26:50
And I think being client centric, it's like a hyper important lesson.

26:50 - 26:55
That's that's completely true, and, and that's the thing, the next thing. So yeah.

26:55 - 26:57
So going especially to that topic.

26:57 - 27:00
So when you get the feedback, usually, it's a it's a data. Right?

27:01 - 27:02
And we are in the era of the data.

27:03 - 27:10
So it's like I can I can I can go in multiple, like, if we are talking about AI right now, what

27:10 - 27:13
AI needs is actually 2 things so you can build an AI?

27:13 - 27:17
You need, a model and you need data.

27:17 - 27:28
And that's for sure if you don't have good data and and then usually Souf use data to to to optimize your Khaled.

27:28 - 27:30
And and that's and that's it.

27:30 - 27:36
Actually, this is the thing that maybe many people, they don't understand how actually AI is behaving.

27:36 - 27:41
AI is actually, a computing probability system.

27:41 - 27:46
So what they have is like when you ask something, especially in ChargeGPT, for example, when

27:46 - 27:47
you say, how are you?

27:47 - 27:50
He gonna check and say for the how are Souf?

27:50 - 27:58
What are the the the the the most probable, let's say, answer that he can give for that?

27:58 - 28:04
It's not really as, smart as people they think. It's actually doing that.

28:05 - 28:12
And people sometimes they tend to fix it in some in some places because sometimes he he doesn't say the right thing.

28:12 - 28:13
And this is the thing.

28:14 - 28:16
AI doesn't give you the right answer.

28:16 - 28:20
It give you the most possible 1.

28:20 - 28:22
And this is this is really very important.

28:23 - 28:26
So I'm talking about the AI right now.

28:26 - 28:29
I don't know in 10 years or in 5 years how it will work.

28:29 - 28:36
This is 1 thing that I understood and actually, like, it make my life easier because sometimes I don't trust it.

28:36 - 28:42
I like when I, when I, when I'm working with ChargeGPT, sometimes it says something that's completely wrong.

28:42 - 28:46
And I tell them, I tell SureGPT, hey, this is wrong, actually, what what you told.

28:46 - 28:48
And you say, oh, I'm sorry and all that.

28:49 - 28:54
So this is thing that you need to to understand about the the limits of AI.

28:54 - 28:58
But we are in the era of data.

28:58 - 29:05
If you have data, a prorated 1, it's something that can be used for different ways. 1 of them

29:05 - 29:08
is actually collecting feedback is going to be as formless data.

29:09 - 29:16
And, and the other thing that also I've seen in data, right now, we are starting seeing data as a product.

29:17 - 29:24
Like, what I'm trying to say is, like, usually when you build application, now is actually most

29:24 - 29:29
of them, they offer also an API or something to expose their data.

29:29 - 29:32
So it can be used by the internal system that you have.

29:32 - 29:36
So you can build something from that. It can be AI. It can be, statics.

29:37 - 29:40
It can be like like like I told Souf, about the traffic.

29:40 - 29:45
Like, how do Souf know that the traffic is actually in Black Friday and Cyber Monday is actually 500%.

29:45 - 29:46
We have data on that.

29:47 - 29:49
So everything is based on data.

29:49 - 29:56
And data is your feedback, especially when you work with, with these kind of big application

29:56 - 30:01
b to c and you have, like, it's not easy to talk directly to the customers.

30:01 - 30:05
You have 1, 000, 000. It it's not easy. You know?

30:05 - 30:10
You gotta take maybe a few few of them, 100, 200, but that's all.

30:10 - 30:12
But more than that, you're gonna need data on that.

30:13 - 30:19
The other also thing that I've seen is, like, we we we talk about open data.

30:19 - 30:25
Sometimes you have data that you can exploit for yourself to use it to build some kind of product on that.

30:25 - 30:28
So this is where we live.

30:28 - 30:32
So the next thing I think is like AI.

30:32 - 30:37
We're going to have AI with more operated data.

30:38 - 30:43
The only issue that I see right now, and I think is going to be solved Souf.

30:43 - 30:46
And that's why actually I will I will explain a lot of stuff that happening.

30:46 - 30:50
The issue with AI right now is actually taking a lot of computing.

30:50 - 30:53
So you need a lot of resources.

30:54 - 31:00
And that's why NVIDIA action is like they they they skyrocket the

31:00 - 31:09
I read today that the value of NVIDIA, it's, like the it's more today than the whole, stock

31:09 - 31:11
market in Germany right now.

31:11 - 31:21
Yeah. Because of AI. And if you hear also their CEO or their CTO talking, they are pushing for AI. Why?

31:21 - 31:23
Because their business is actually based on that. Yeah.

31:23 - 31:26
So and that's completely, completely normal.

31:26 - 31:33
And even now I actually I heard that Elon Musk has actually started building his company, the,

31:33 - 31:38
the XCI, who is actually trying to to be a concurrent to to open AI.

31:38 - 31:44
And, first thing is actually big contract with NVIDIA. So yeah.

31:45 - 31:50
And it's the same also for cloud providers like Amazon or like like Google.

31:51 - 31:55
First thing, when they start to build an AI, they they they ask for resources.

31:55 - 31:58
And today, the biggest 1 is actually NVIDIA.

31:58 - 32:04
So They are so so lucky because, like, during the gold rush, the

32:04 - 32:10
the the people that make most money out of the gold rush are the people that's selling selling the shovels. Right? Yeah.

32:10 - 32:12
Not who are digging the gold.

32:12 - 32:16
And before, they had the the Bitcoin and the miners. Right?

32:17 - 32:23
And now they have the AI, which is, like and but it's unbelievable how how their their company is going.

32:23 - 32:28
Yeah. So you you bring it in, like, a like, a very interesting subject that I'd like to go to.

32:28 - 32:34
It's like, in my opinion, usually when you have a technology that is starting, you need to wait

32:35 - 32:39
until it matures the usability or the usage of that.

32:40 - 32:46
And then from there, you, you can, like, you can, you can start benefit from it.

32:46 - 32:55
What I'm trying to say is, like, if you go to blockchain so the first thing is actually cryptocurrency than NFT.

32:57 - 33:00
In my opinion, these are only the first steps.

33:00 - 33:07
We're gonna have maybe in the future, we're gonna have a third 1 or maybe a 4th 1 use case that

33:07 - 33:12
is going to be more, let's say benefits than these first cases.

33:12 - 33:15
These 1 are only the first cases of, of blockchain.

33:16 - 33:17
Why I'm trying to say that?

33:17 - 33:22
Because usually the first use case and the second 1, they they are killed at some point.

33:22 - 33:24
But maybe it's it's still my opinion.

33:24 - 33:29
But, we we will see in the future how how it happens. Why I'm saying that?

33:29 - 33:31
Because it was the same with microservices.

33:32 - 33:38
Microservice is actually the first technology that is built on that was called Vertex.

33:39 - 33:44
Vertex is actually like, nobody is using it right now.

33:44 - 33:50
But now people, they are using the tools that were built by Netflix, for example, in the the

33:50 - 33:55
Java stack, like, like, Erica, like, like the circuit breaker. I don't know.

33:55 - 33:58
I forgot the the the names of the of the tools.

33:58 - 34:06
But what I'm trying to say, when a concept happens and is created, the first user is actually

34:06 - 34:09
most of the case, it will not survive survive a lot.

34:09 - 34:17
You need to wait 2 to 3 to 4 years until until we have, like, other cases, other use cases.

34:18 - 34:20
And these, most of the case, they're gonna survive.

34:20 - 34:22
The AI is the same.

34:22 - 34:29
We got the hype right now, but at some point, it's gonna be converging to very specific user.

34:29 - 34:36
1 of the users that that I see right now is, like, Google is actually, they make a lot of money

34:36 - 34:38
from the, the search engine. Right?

34:39 - 34:47
I think part of that market or maybe even the whole market of that gonna be switch it to to to the AI.

34:47 - 34:53
Because right now, the people when they do a search in Google, they search for specific information.

34:53 - 35:01
And usually that information is actually not treated, you know, not, like, you need to process

35:01 - 35:05
it as a as a person and then take what it should be taken.

35:06 - 35:16
That thing is actually already done when when you ask a shot GPT or you ask any any any AI that is exist existing. It's still not perfect. That's for sure.

35:16 - 35:18
But we are going there.

35:18 - 35:22
Gonna take some time, but we are going there. Yeah.

35:22 - 35:24
So we switch it to the AI. Sorry about that.

35:24 - 35:28
No. No. That's a that's a really interesting, topic.

35:28 - 35:33
I don't know if you know this book Crossing the Chasm, like adaptation of the users.

35:33 - 35:36
They're like Souf have the innovators at the very beginning. Right?

35:36 - 35:39
And then it's going up, like, the the the chart.

35:40 - 35:43
So III think this is the similar case.

35:43 - 35:49
It reminds me when I was starting the company, the the the the Brain Hub, the the agency that

35:49 - 35:54
I have, like, in 2, 000 16, we started as a JavaScript, like, best JavaScript team. Right?

35:54 - 35:57
And JavaScript back then, it was just starting and growing.

35:57 - 36:02
We set the bet on React, on Node, and it was just a start. Right?

36:02 - 36:06
And when the with the adaptation was higher and the libraries frameworks, they were going, like,

36:06 - 36:10
having more users and bigger compar corporations are using it.

36:10 - 36:12
So there were more tools, more demand. Right?

36:13 - 36:18
But at the same time, as I don't know if you remember, but number of frameworks and libraries

36:18 - 36:23
that were competitors to those technologies and everything, like, with the JS at the very end,

36:23 - 36:26
like, each day Souf have, like, a 3 different new things. Right?

36:26 - 36:31
And the developers, they want to build the stuff with the new things because this is exciting.

36:31 - 36:37
But at the same time, if you are, working with the company which has, established which is established

36:37 - 36:43
on the market, you cannot, like, go and say, like, all in into new concepts. Right?

36:43 - 36:48
Like, in a completely new tools, libraries because you cannot predict what will happen if this

36:48 - 36:50
will survive in 2 years. Right?

36:52 - 36:59
It's in the market. So it's like, we people tend to forget that we are in, what you call, a new science.

36:59 - 37:01
It's still a new science.

37:01 - 37:03
The the tech industry is still a new science.

37:03 - 37:09
So it's not like mechanical engineering or, like, let's say civil engineering or something,

37:10 - 37:14
like like, who has actually been there for 1000 of years. You know?

37:14 - 37:17
We've been there, like, maybe for 50 years or something.

37:17 - 37:22
So we we we we tend to to it's going fast.

37:23 - 37:27
It's gonna stabilize at some point. That's for sure. I hope not.

37:27 - 37:32
Because if it if it's still there, we still have work to do. But yeah.

37:33 - 37:36
That's the the thing that, let's say, exciting and exhausting in the same way, in the same place, Souf know,

37:44 - 37:46
there is always new stuff that you need to learn.

37:46 - 37:49
And what is exhausting is, like, you need to do that all the time.

37:49 - 37:51
It's like people, they are reading newspaper.

37:52 - 37:58
I think most of the techies, me including myself, usually our newspaper is, like, looking into

37:58 - 38:00
what happens in the tech today.

38:00 - 38:02
What what which startup is actually going up.

38:02 - 38:04
Where where we are going?

38:04 - 38:07
What are the the techniques that we use right now?

38:08 - 38:13
What that company do we actually successfully use as a process and how they build that process.

38:14 - 38:16
And we always tend to do that.

38:16 - 38:19
And this is actually also big issue for most of CTOs.

38:19 - 38:22
So how how do we keep up with all of that?

38:22 - 38:26
Well, 1 thing that I do is like, I can't keep up with everything myself.

38:27 - 38:32
So what I need to do is like, I keep up with with which I can keep keep up.

38:32 - 38:35
And then I share it with some other people who is actually in the same state as me.

38:35 - 38:37
And the same goes for for others.

38:37 - 38:43
Then when you have that network, this is what I call well, what I call maybe what most of the

38:43 - 38:46
people call collective, intelligence or something.

38:46 - 38:54
So so it's like, you you bring a lot of people even if, like, not all of them are smart or maybe

38:55 - 39:00
not nobody is actually smart when they are together, they are smart actually because they are

39:00 - 39:04
sharing knowledge, they are sharing techniques, they try different things.

39:04 - 39:10
And then, so so that's why when when you are talking about CTOs, it's very, very important to

39:10 - 39:13
be in a community of CTOs or maybe a community of techies.

39:13 - 39:15
So you can keep up with that.

39:15 - 39:20
And you can, like, check how how to do things, how happens.

39:20 - 39:26
And then, and that's actually pretty cool in tech because most of people that they are in tech,

39:26 - 39:35
they are, by default, in my opinion, they are people who tend to share a lot what they discover. And that's good, actually.

39:35 - 39:38
I hope I hope it's the same in the other this.

39:39 - 39:45
I I think, like, exchanging knowledge is 1 thing, but for me as a cofounder, I talk with a lot

39:45 - 39:48
with my competitive stores from all around the world.

39:48 - 39:51
All, like, the leaders from the disciplines that I'm interested in. Right?

39:51 - 39:58
But I'm doing it not only to exchange the knowledge, but even for the mental health because,

39:58 - 40:00
like, as a leader, you are alone.

40:00 - 40:03
Like, you have a feeling that you are alone and you are alone with the problem.

40:03 - 40:08
But apparently, when you talk with other leaders, they have the same problem at the same time.

40:08 - 40:12
So it's, like, even good for the mental health to be in touch with them.

40:12 - 40:15
Because like you said, it's very good for the mental health.

40:15 - 40:21
And, also, you figured out that you are not the only 1 facing these kind of problems.

40:21 - 40:26
And that's actually pretty cool because everybody has a way to solve them.

40:26 - 40:32
And you can from that, you can develop your own way based on the experience of the others. 1

40:32 - 40:38
other thing that that I tend to say is, like, some people, like I told you about the 10 years

40:38 - 40:43
of experience and between 1 multiplied by 10 and the 10 years, usually the people that have

40:43 - 40:47
10 years of experience, when I say it's like, it's a figure of speech, it's like can be 10,

40:47 - 40:58
20, 30, They tend to have people around them in the same mindset so they can keep up about what happens in the industry.

40:58 - 41:01
So they have 10 years of experience.

41:02 - 41:09
But in reality, with all the accumulated knowledge that they have, they tend to behave like

41:09 - 41:16
somebody who has 30 years of experience because of all the experience that they they they accumulated from the others.

41:16 - 41:22
And the same also when I say that, the same also goes for their peers because it's the same.

41:22 - 41:25
They they are doing the same, the same stuff.

41:25 - 41:31
And that's actually pretty cool because you can, especially and this is an advice for junior

41:31 - 41:33
developer, for junior tech tech.

41:34 - 41:41
If you are, in these kind of communities, you're gonna grow faster than somebody who is actually

41:41 - 41:43
is not in this kind of communities.

41:43 - 41:50
And these people tend to to go, like, to to to to be successful in what they do.

41:51 - 41:54
And this is actually 1 of the questions that I remember, how to be successful.

41:54 - 42:02
I think you need to be round people who is actually positive, the same mindset, and working,

42:03 - 42:08
in the same field, like like like like I told about the mental health and other stuff.

42:08 - 42:16
And in the same way, you can you can have multiple vision about the same, you know, this glass

42:16 - 42:18
this is a glass of water.

42:18 - 42:24
Somebody gonna say the full, to to see the full, the full glass of water.

42:24 - 42:28
Somebody else is gonna say the empty, that that is in the glass.

42:28 - 42:31
So and it's very interesting because it's 2 visions.

42:31 - 42:34
So you need to understand there is other vision than yourself.

42:35 - 42:37
I have a similar concept.

42:38 - 42:41
I I call it, like, outsourcing of decision making.

42:41 - 42:46
So I always think, like, I have my opinion, but I always before, like, making the decision that

42:46 - 42:53
impacts, like, a lot of people or maybe, you know, the way we way we do things, It's like, I

42:53 - 43:00
have, like, a group of a few people, but they are sometimes from different fields regarding entrepreneurship on different levels.

43:01 - 43:04
They are even out like, they yeah. Outside of tech. Right?

43:04 - 43:11
But I really respect and like their opinions and how they think, and I crushed my decision making with them.

43:11 - 43:15
Even once, I had that psychotherapist. Right?

43:15 - 43:20
And I went to her, like, and I made we made, like, a 4 session, and she helped me a lot.

43:20 - 43:24
And I was, like, amazed because she was just listening what I'm saying.

43:24 - 43:29
I'm talking about the business, how I structure the company, how I structure the teams, like, how do we work.

43:29 - 43:33
And she's you know, she has no clue about doing the business.

43:34 - 43:38
But at the same time, she's listening to what I'm saying, and it's all a a lot about working

43:38 - 43:44
with people and how to approach and your mindset that you don't see from the outside how you're making the decision making.

43:44 - 43:51
So she helped me to structurize the the setup of the team, like, completely outside of the box.

43:51 - 43:53
The same when I'm talking with the guests of the podcast.

43:53 - 43:58
A lot of them are saying, like because I'm asking about that, and they were asking the same question.

43:58 - 44:07
There'll be others about, like, some books or, resources, conferences that have been particularly influential on, on that. Right?

44:07 - 44:09
What they have learned all of it.

44:09 - 44:12
And a lot of them start with the centers like, hey.

44:12 - 44:18
Maybe this will surprise you, but the book that influenced me, it's not about the tech. Right? It's about the stoics.

44:19 - 44:23
It's about, like, Mark Auralius. It's, like, philosophy.

44:23 - 44:27
Like, something but but this has influence on Nelson.

44:27 - 44:33
Yes. True. This is actually quite interesting. 1 of the book that, like like, it changed my

44:33 - 44:42
perspective on how, you talk to people is actually, I think it's called how how to how to win friends or something.

44:43 - 44:45
How to, how to make friends and influence people.

44:45 - 44:47
Daniel Carden. 1 of my friends.

44:47 - 44:51
I read it, like, 5 years ago or maybe 6 years ago.

44:51 - 44:57
I still have a copy in, at home, but that thing's actually changed my perspective about how

44:57 - 44:59
how to to deal with people especially.

45:00 - 45:03
And this is 1 of the book.

45:03 - 45:06
The the other book is actually called, thinking in systems.

45:06 - 45:12
Because when you are dealing with people, people are systems as, like, especially when you're

45:12 - 45:14
dealing with teams and you are a leader.

45:15 - 45:20
You need to to think in in matter of systems and how the system, they influence each other.

45:22 - 45:23
I can give you a third book.

45:23 - 45:26
It's actually 1 of the best seller, and I like it a lot, who is actually

45:29 - 45:32
Thinking Fast and Slow. I think they know about the book. Yeah.

45:32 - 45:38
That that book is actually incredible about how how they did the experience and how the how

45:38 - 45:39
they found about the 2 systems.

45:40 - 45:46
The system was actually lazy, but in the same time, very pragmatic and very, like, the 1 who

45:46 - 45:48
is actually thinking, let's say.

45:49 - 45:53
And the other who is actually very fast, but in the same time, it's intuitive.

45:53 - 45:58
And how these 2 systems inside your mind, they are influencing, each other.

45:59 - 46:01
And it's really very, very interesting book.

46:01 - 46:05
I think I recommend it for any person not only techies.

46:05 - 46:12
Any person should read that book because there is a lot of insight in that book that can be taken in real life.

46:13 - 46:21
The other thing that usually when I tend to when somebody is asking about something, when people

46:21 - 46:26
try to understand the concept, don't ask people.

46:26 - 46:27
Go to the founder of the concept.

46:29 - 46:34
Buy his book or read his blog Souf he can get the essence of that book. Why?

46:34 - 46:42
Because always in my head, when somebody try to read about something, about the the 1 person

46:42 - 46:45
that didn't, let's say, create that.

46:45 - 46:53
It's like asking somebody to about, let's say a movie that you you've seen. Go see the movie.

46:53 - 47:00
And then from there, you can ask other people about their opinion so you can see different.

47:00 - 47:04
But go see the movie so you can get this as yourself.

47:04 - 47:11
And then from there, you can you can start, like, discussing with others to see how they they've seen the movie.

47:11 - 47:14
What What are the parts that maybe you didn't see about the movie?

47:15 - 47:21
What are their perspective about the cast, about the actors, about the

47:24 - 47:29
the sharkter. So and there, it's becoming interesting conversation because you have already

47:29 - 47:32
your opinion and you add to them.

47:33 - 47:38
Going back to making decision, I like I like to go back to that.

47:38 - 47:48
So 1 of the things that I learned the hard way is, like, when you are a leader or you become a leader,

47:51 - 48:01
talk less, listen more, and never take a decision before hearing all the people around you that

48:01 - 48:03
gonna be impacted by that decision.

48:03 - 48:08
What What I'm trying to say is like, even if Souf know for sure for that, you're going to make

48:08 - 48:12
that decision, give them space so they can share their thoughts.

48:12 - 48:15
Because maybe there is some few things that you don't know.

48:15 - 48:19
And in the same time, you give power to your team actually.

48:19 - 48:23
When you do that, they they know that their opinion is valued.

48:24 - 48:34
And they know for sure that even if you take the decision, you take it based also in their, input. This is 1 thing.

48:34 - 48:41
The other thing, if there is, like, a conflict, when I say, like, for taking decision, don't worry about that.

48:41 - 48:43
Pick up 1 of them.

48:43 - 48:51
But always make sure that that decision or maybe most of this decision that you take are reversible.

48:52 - 49:00
So what I'm trying to say is, like, a decision can be good or bad and can also be good or bad

49:00 - 49:05
depending on that context, on that specific, let's say, day.

49:06 - 49:13
But in a few years, maybe that decision gonna be wrong because the context change it and a lot of stuff have changed.

49:14 - 49:16
In some places, you need to take decision.

49:16 - 49:21
So take the decision by but make sure that it can be reversible.

49:21 - 49:26
So what is interesting is creating a process that can reverse the decision.

49:26 - 49:31
So that's why most of the people they focus they tend to focus on learning curve because they

49:31 - 49:35
say that, how can I go and do that Souf?

49:35 - 49:41
I'm getting a faster feedback, so I can switch my decision if if needed to be.

49:41 - 49:45
And this is actually the principle of reversible decision.

49:45 - 49:51
So we need always to have as much as possible reverse re reversible decision. Why?

49:51 - 49:56
Because at some point, you can take you need to take a decision. Take the wrong 1. That's okay.

49:56 - 50:04
But you can then then make it make it reversible so you can change it in in in any days after that.

50:05 - 50:09
I think Amazon popularized this concept, right, about, that that I think yes.

50:09 - 50:12
Is this decision or service irreversible?

50:12 - 50:17
So, yeah, a lot of amazing tips, Khaled.

50:18 - 50:25
And you mentioned the thing about, the first 1, yeah, about the meetings and, like, being silent on the meeting.

50:25 - 50:28
So this was the lesson I learned 2 or 3 years ago.

50:28 - 50:30
So as a leader, you want to delegate.

50:30 - 50:35
You want to have the leaders, and you want the people inside the organization to communicate and make the decisions. Right?

50:35 - 50:35
Mhmm.

50:35 - 50:39
But as a cofounder, it's very tempting because you know the organisation.

50:39 - 50:40
You are thinking really fast.

50:41 - 50:47
You made a lot of decisions, so probably you could make the decision 10 times faster because you know more context. Right?

50:48 - 50:53
But, in this way, if you are making the decision, then the company is dependent on you so much,

50:53 - 50:55
right, that you want to create the system.

50:55 - 50:58
So I had the rule I that I'm going to the meeting and I shut up.

50:58 - 51:06
So I I don't say anymore, And after after a while, like, people have started to kick me out

51:06 - 51:09
of the meetings, out of the channels on the on the Slack. Right?

51:09 - 51:11
Because you are not needed anymore.

51:12 - 51:17
And I think this is the best that this is the goal as a leader that you want to get. Right? Exactly.

51:18 - 51:25
Also, there there is also another aspect about that why you should, because when you go up in

51:25 - 51:29
the hierarchy, your words have more power.

51:30 - 51:33
So you need be very careful when you use them.

51:33 - 51:39
Like, what I'm trying to say is, like, when you are talking to a coworker, you can talk the

51:39 - 51:47
same way to somebody who is actually, like, when you are a leader of the team because, there

51:47 - 51:48
is there is a barrier.

51:48 - 51:55
And in the same time, sometimes when you when you argue about something, he can't answer you

51:55 - 51:56
back because of the hierarchy.

51:56 - 52:00
You need to create some safe space for people to do that.

52:00 - 52:08
So it's like, never, let's say, don't tell people to do something, ask them to do.

52:08 - 52:17
And make it for them Souf even if you ask them to do that, they know that for sure they can

52:17 - 52:19
they can argue about your decision.

52:19 - 52:25
And that's very cool because when people really trust you, they're gonna start arguing about,

52:25 - 52:28
like, challenging you for the decision. And that's cool.

52:28 - 52:34
If if they are challenging you, it means that they trust you and they care about the product,

52:34 - 52:36
and they care about the company.

52:36 - 52:44
And this this is really very, very important to create as a space, as a culture of a company.

52:44 - 52:51
So you can be sure that, like you said, is going to be a system that is working by itself without Souf, influencing it.

52:51 - 52:58
Influencing it may be in some places, but but it will still work if you take, like, 1 month

52:58 - 53:02
of vacation or something because everybody knows what it need to be done.

53:03 - 53:07
I remember that somebody done that metaphor about, founders.

53:07 - 53:16
He say, the if we if the company is a car, the founder should be, the 1 who is actually switching,

53:17 - 53:22
like, let's say, pieces if there is a piece that that is not working. And that's all.

53:22 - 53:24
Because the car is already working by itself.

53:25 - 53:32
And and what is gonna be true more and more with Tesla right now with the system with the AI system, we hope so.

53:33 - 53:38
But, yeah, you you get my point about the the the metaphor, actually, about that.

53:38 - 53:42
So building systems that work by themselves, that's actually very interesting.

53:42 - 53:49
And I remember actually also 1 of the cofounders, he said that attaining some number of employees,

53:49 - 53:57
well, the CEO doesn't know what what to do anymore because everything is actually working by by itself. So

53:59 - 54:06
I would just recall the 1 thing that you said about the Daniel, that's how to win first and influence people book.

54:06 - 54:11
This is interesting regarding the operating system, how, like, we deal between the people because

54:11 - 54:17
this book is, I feel, like, 70 or 60 years old, and it's still valid because it's still the

54:17 - 54:25
way how the people interact with each other, make the decisions, get angry, like somebody, it's

54:25 - 54:27
like the same kind of, thing.

54:27 - 54:32
So it didn't change so much, and you still can apply it wherever you are.

54:32 - 54:36
But, the the last question that I wanted to ask you because, like, we talk about a lot about

54:36 - 54:41
the experiences that you already have, but we are constantly learning as a people.

54:41 - 54:45
So on the tech part, I assume you always follow, like, the trends and that kind of Souf.

54:46 - 54:52
But are there any pain points, challenges on which you work currently or you educate yourself?

54:53 - 55:01
Well, there is, what I what I say is, like, what I see is, like, where there is a pain point

55:01 - 55:07
or let's say, when you are outside of your comfort zone, you are, in the learning zone.

55:07 - 55:08
So that's a good thing.

55:08 - 55:15
Even if it's not really very comfortable sometimes, but for sure, you are learning something.

55:15 - 55:24
But if you stay in the same space with the same stuff, doing the same things, well, let's say

55:24 - 55:27
an AI can can replace you at some point

55:29 - 55:33
right now. So, the idea is, like, yes, there there is pain point.

55:34 - 55:41
Souf, for example, I can I can I can give you 1 1 of the issue that I have, like, few weeks back?

55:42 - 55:49
It's like, I'm discovering right now how to to reach out customer by myself because usually

55:49 - 55:51
I delegate that to somebody else.

55:51 - 55:56
And, it's not easy, but I'm discovering the world of the sales.

55:56 - 56:01
But now I have I have some kind of sales with actually coaching me how we should approach customer,

56:01 - 56:03
how we should, like, talk to them.

56:04 - 56:11
When I say customers, I mean, in very specific area because when my customer is a techie, it's easy.

56:11 - 56:15
When your customer is something else, it's another word. Right?

56:16 - 56:23
So you need to to figure it out how to approach them. And it's not easy.

56:23 - 56:28
These are actually was like, but now it's actually it's becoming less of a pain point because

56:29 - 56:37
and and I need to figure out something else so I can so I can learn something new, you know. So, yeah.

56:37 - 56:41
What I'm trying to say is, like and this is the advice that I give mostly to techies.

56:44 - 56:50
Let's, let's remove the image that most of the people they have about us techies.

56:50 - 56:56
Somebody who's actually a nerd, staying alone, doesn't talk to people, always playing games

56:56 - 57:04
after coding, or maybe always being, being at home, and always having delivery food and all

57:04 - 57:07
the staff doesn't go out. It's not true.

57:07 - 57:12
The techies of today, they should be more versatile than that.

57:12 - 57:20
And our work is not actually about only the code is actually bringing value to people. That's a big difference.

57:21 - 57:27
So if you are only building code, doesn't make sense that you are not, like, contributing to

57:27 - 57:29
to the society as as a person.

57:31 - 57:39
But taking that and, let's say, sharing with others, And especially if you are a senior leader,

57:39 - 57:44
it's like even better because you can teach people and you can mentor people on that.

57:44 - 57:48
So it's like you can bring more value than you think. Love that.

57:49 - 57:55
And this is actually the best that you can do is like doing, creating company, managing company,

57:55 - 57:58
and then maybe mentoring people to do that.

57:58 - 58:06
Mentoring people to to to take executive level, like, suite for for the future.

58:06 - 58:11
Do it like like like like Souf at some point, the community helped you.

58:11 - 58:13
So why not give back to to the community?

58:13 - 58:16
And I'm not talking about only the tech communities. It can

58:16 - 58:23
be also any type of community that helped you to do what you do today. Awesome.

58:23 - 58:24
Thank you very much, Khaled.

58:24 - 58:27
It was extremely insightful talk.

58:27 - 58:30
For the first time, I haven't used the questions that I had inquired.

58:31 - 58:40
So it was, a lot of interesting topic that we could probably continue, over, over another hour. But, thanks for that.

58:40 - 58:43
You're you're welcome. Thank you for having me today.

58:44 - 58:49
Follow Matt on LinkedIn and subscribe to the Better Tech Leadership newsletter.

‍