‘Tech teams: Assess, Adapt, Transform!’ with Andrei Gavrila, Agile Coach & Fractional CTO at Pentalog

Kovid Batra: Hi everyone! This is Kovid, back with another episode of Beyond the Code by Typo. Today we have a special guest from Romania with us. He has 16+ years of engineering and leadership experience. He’s working as a fractional CTO with Pentalog. He believes in consulting, and not just consulting, but then implementing solutions with the clients. He does technical audits, MNAs for all the tech teams. And he’s one of the highly recommended mentors on MentorCruise and Plato. Welcome to the show, Andrei. Happy to have you here. 

Andrei Gavrila: Happy to be here too. 

Kovid Batra: Great. Thank you so much for taking out time. And before I get started and our audience gets a lot of learning advice from you, I would like to know a little bit more about you, the audience would like to know a little bit more about you and we can, like start off with something like, how do you, how do you unwind your day? Where do you spend most of your time when you’re outside of work? 

Andrei Gavrila: Good. So, I like to spend time with my family, so that would be number one. Then, I relax through learning. Maybe that sounds funny, but not to me. I really enjoy learning. So, usually at the end of the day, I am trying to squeeze one, one hour, maybe half an hour, so that I learn new things and I also like to play video games. 

Kovid Batra: Oh, nice. Video games. So, this is like a childhood thing and is it like something which started very early on?

Andrei Gavrila: Yeah. Yeah. I started playing, I think, video games when I was five years old. 

Kovid Batra: Oh yeah! 

Andrei Gavrila: I don’t even remember what the name of that computer was, but the first game that I played was called ‘Prince of Persia’. 

Kovid Batra: I remember it completely, man. All right. Which one are you playing these days? 

Andrei Gavrila: Nowadays I’m playing a game that’s called ‘Total War: WARHAMMER’. It’s a strategy game. I think it works really well with the role that I’m doing that has a very strategic part. So yeah, I enjoy strategy games. 

Kovid Batra: Perfect. Perfect. That’s an interesting find about you. All right. Apart from like unwinding your day with games and maybe taking some learning from there, what are the other learning sources that you have in your life? Like, do you do, read books or watch videos, listen to podcasts? 

Andrei Gavrila: So, I’m trying everything, but maybe what the best learning tip that you might get from me is that not necessarily sources, but the way that I learn. So to me, first of all, learning is not reading or watching videos. To me, learning means changing behaviors. When, when we do that, the learning cycle, in my opinion, ends. And I think a lot of us associate learning with maybe reading or watching videos, but I think until the moment that your behavior doesn’t change or you don’t change anything, uh, I wouldn’t call that learning. And one of the things that I do to learn is I use a framework that’s called the 70/20/10 that says that for every, let’s say 100 hours, you should try to learn 70 of them through experience. So your job, the regular day-to-day stuff, 20 of them by learning from peers. Even what we do now, I would call it engagement learning from peers. So, it doesn’t necessarily need to be too formal. And 10 of these hours, learning in a structured way. So things like books, videos, podcasts, exactly what you said.

Now, if I were to recommend some of the let’s say some of my learning sources, there are many. I will share with you and maybe the audience, Trello board that I have where I’m keeping track of the learning that I do, but most of all, I would like to say that since I’m using ChatGPT many times a day, I feel that that boosted my learning significantly. Sometimes I even do let’s say, interviewing styles with ChatGPT where I’m asking it to ask me questions about a specific topic or a book I might have read. I think that’s really, really amazing. 

Kovid Batra: Oh, that’s really interesting. We have always been asking questions to ChatGPT, but we never asked ChatGPT to ask us questions so that we are able to learn more. That’s, that’s very interesting. I would love to see that Trello board also. So whenever this podcast is going out, I’m sure, in our comment section, we would love to see that coming from you. 

Andrei Gavrila: Sure. Definitely. 

Kovid Batra: Amazing, Andrei. And, thank you for sharing your personal way of learning and some thoughts on that.

Let’s move on to the main section, uh, where we would love to know what you do in day-to-day as a fractional CTO, like things like how you’re managing your teams there. First of all, I just wanted to understand, when you work as a fractional CTO, right, are you specifically working with small-size teams or is that these large-scale organizations? What kind of organizations do you work with? 

Andrei Gavrila: Yeah. The term ‘fractional’ can be confusing. Sometimes it might be confused with just ‘consulting’, sometimes it might be confused with ‘interim’. We will not get in the depth of the term because I don’t think it’s that interesting, but I like to work with all types of companies. Most of my experience is around, let’s say medium to larger companies. In reality, the term ‘fractional CTO’, I think works a bit better for startups. For example, you can be a CTO for two to three startups and you’re fractional there, fractional there, fractional there. But me, my, the things that I enjoy most doing is work with bigger companies that have problems around scaling, that see opportunities around these economies of scales delivering to millions of users that are looking for more efficiency, that are looking for predictability, they try to mix a world of legacy with the innovation that we can do today. And they are working a very fine balance between, let’s say those two domains. 

Kovid Batra: Makes sense. Perfect. So, when you say it makes more sense with startups, I think you must have had some experiences with startups and you must have seen the challenges of being a fractional CTO with these startups, right? And I’m sure if you, if you got a chance to work with large-scale organizations, there, there would be a different set of challenges. So let’s start with the startups first, like what kind of challenges do you see and how do you tackle them, and then maybe we can discuss about something, for large-scale organizations, how to effectively build teams there and, like execute with teams there. So.. 

Andrei Gavrila: Sure. I would tell you first that in Pentalog, we define a CTO assessment. So, because we work with so many CTOs and we see the role implemented differently, we said we would like to help people better understand what the role of the CTO is, how it should be done and what different implementations of this role we see in the industry. And because of that, we define services around that. One service is CTO assessment, helping CTOs understand how they’re performing their roles. This is used both for growing or for recruitment. Then we have services around mentoring and a lot of the consulting that we do, also helps this role. But the idea is that in general, what we see is that in the startup world, the role of a CTO is more, is closer to a technical lead role, right? So, somebody who is coding and driving the technology very hands-on. So, sometimes the startup CTO is the person that has the most development experience, somebody who will code the hardest parts. And that’s normal because that’s what startups do. They try to put a product in our hands as quick as possible. You want to create the product. But, they are in a way exposed to the risk of missing the strategic part of the role of the CTO. So things like governance, things like how do we make sure that the decisions that we take today are still in control and will still be in control, once we succeed, once we want to scale, once we reach a phase of hyper-growth. These kinds of things are not really the best things that technical lead or startup CTO could answer. So, that’s the kind of work that I do when I work with a startup CTO. So, bringing more of that strategic governance, efficiency, decision, cost of ownership review. 

Kovid Batra: Great. Perfect. But don’t you think, like with startups, which are in early stage figuring out product-market fit the strategic thinking could not be very, like long-term, right? Because they don’t know what they are going to build. Sometimes it happens, the customer itself changes, right? You shift from one segment to another. So, in that scenario where there are so many uncertainties involved, maybe you yourself would want to keep things streamlined in terms of strategy, but how do you communicate that and create a balance with that dynamic environment? 

Andrei Gavrila: Yeah, that’s a, it’s a really good question. It’s not about bringing the kind of order that you would find in medium or larger enterprises, but it’s about finding the right balance as you say. I would give you an example. Startups usually decide without really analyzing and really understanding their decisions because they are in this space where they need to take a lot of decisions and there are some decisions that can really impact the long run. And also there are some decisions that become a problem after a while. One of the things that I do when working with startups, I’m saying we can decide whatever we want, but how about we try to keep track of the major decisions, and to explain it to our future self, why we took them in this way, and maybe when this decision will expire, or we need to revise it because something happened. I think this gives a lot of clarity in the life of a startup because it allows it to plan the development. So for example, you would take one decision around hosting or infrastructure or several capacity. But if you don’t specify that that decision will no longer work when you reach 10,000 users, you might reach 10,000 users and face that problem just then. 

On the other hand, if when you decide, you know when that break up point might be, maybe when you go towards the 10, 000, you go close to 6, 000, we’ll start working on that because you have this clarity around what you decided and when these decisions are not working anymore, right? This is the kind of, let’s say, strategical or governance-thinking that is not natural for, uh, as I said, the startup CTOs, because their context is much more now let’s put the product with the right set of features in the hand of our users, but nevertheless, they decide and sometimes these decisions don’t remain in control. 

Kovid Batra: No, absolutely, I get it. And it’s, it’s more about just taking a conscious decision whenever you are moving ahead. And at least you would be prepared, you would take a few better steps to handling the future problems also. So, it makes complete sense, absolutely. 

Other than that, like moving on from startups to some, let’s say large-scale organizations, you mentioned in our last conversation, when we were talking that you take care of the agile, implementation of agile methodologies in the teams and you do that with different companies, right? So when it comes to large organizations where you see these methodologies are being wrongly perceived or people are executing in a wrong way, how do you bring that agile transformation with such large organizations where there is so strong cultural set up already there when you are entering as a fractional CTO? So, is there a good piece of advice for people who are looking to implement agile methodologies at scale? 

Andrei Gavrila: Yes. I hope I got your question right. I will try to answer it and let me know if you, I understood it right. A lot of companies today are in a state of continuous transformation and constantly wanting to do more and better. I think the term ‘agile transformation’ has its benefits. I think it has less and less benefits, because kind of gives this impression that there is a finite thing that needs to happen and then we are transformed somehow. I think the context today is much more one in which this constant adaptability to what happens around us is permanent. And I think we are less in a phase to speak about agile transformation than speak of somehow the new normal, which is, how do I make sure my company is fit so that it can adapt no matter what happens somehow? 

Kovid Batra: Exactly, exactly. 

Andrei Gavrila: And a lot of companies that are maybe trying to work on that now are going through two main ways, especially the big companies. One which is the ones that we see most often where their transformation is very tied to frameworks and methods or topologies or in a way, processes, if you want. So a lot of companies are saying, “Let’s put Scrum in place or let’s put SAFe in place.” So that’s, in my opinion, most of the industry, that’s what most of the industry is doing. And that’s one way to approach that. 

The second way to approach that we also see in the industry is something around culture, something around the way that we look at problems, mindset. A lot of companies will speak about proximity to customer, which is really nice because in a way that’s what agility is. But, a lot of companies are approaching that transformation saying, “Let’s try to change the mindset, the culture in our company to be closer to the customer.” And these are the two big ones to do, the two ways that companies today are approaching transformation and the reality is that to some, I wouldn’t say most, it could be a debate if it’s somewhat most, it’s not really working. They are not seeing benefits. This is why in Pentalog, what we do is we go to a third way which it’s not that obvious in the industry. And this third way, it’s about structure and government. Maybe I’ll give you an example so that everybody understands what I’m trying to say, but our idea is what kind of rules do you need to respect in order for you to be adaptable? 

So, let’s give an example for the these three ways of doing it so that it’s a bit more clear. As I said, for the first one companies would say, “I’m going to be more adaptable because I’m going to use Scrum/SAFe, and this is a structure that allows for adaptability.” So if I’m having the structure, then I will probably adapt. That’s the, the philosophy there. The other one says, “I’m going to try to change people’s mindset, the way they approach things. And they maybe try to go then to a process of training, a process of understanding better, how things should work, speaking more of a customer, placing customers closer to the way that they will, they work.” That’s how they think, it will go. And in our case, we come and say, “What kind of rules do we need to respect?” And one of that rules could be that I’m having cross-functional teams so that I’m all over the organization, I’m having teams that don’t depend on other teams to deliver value. and that’s, for example, one of the, that’s the way that we do transformation, which is in my opinion, working much better than the other two ways.

Kovid Batra: Perfect. I think this is a very relevant example and the thoughts that teams should look into and probably when, when we talk of agile, we probably always look at speed, implementing certain practices like scrum, but the core fundamentals that need to be implemented are around customer-centricity and bringing adaptability with the change. And I think one of the ways that you suggested right now perfectly fits in. 

So, cool. This was, this was really nice. But, there is one more question, like quite unrelated, but I feel that when you’re working with large organizations, you divide them, you decide the right structure, topology for the team you’re working, But, at that such large scale, one thing that goes missing is how do you look into whether the teams are succeeding or not, right? Engineering teams are succeeding in what they want to do or not. So, what kind of framework or methods or let’s say the metrics you would use to understand that? That whether these teams are succeeding or not succeeding, how do you define success for them? 

Andrei Gavrila: Yeah. I’m really passionate about that question and I think we could talk about it for hours. I will try to answer it maybe in the next two to three to four minutes. 

Kovid Batra: Sure. 

Andrei Gavrila: The thing that’s most important to me is that a lot of companies are asking that question, how do I know if my teams are efficient? How do I know if my teams manage to do what, what we need them to do? And the approach to that is usually to define KPIs and see if the teams reach those KPIs or not. Some companies look at velocity. I don’t recommend doing that. Some companies look at output. Some companies look at impact, outcome, that’s much better than velocity, outcome and impact. But to us in Pentalog and to me, the question is not, how do I know if my teams perform? Right? The question that I like to ask first is how do I know that my teams are in a context where they can perform efficiently, where they can be excellent? So instead of asking, “Is this team performant?” I like to ask the question, ” Is this team in a context that allows for performance? 

Kovid Batra: Makes sense. Perfect. I think that that’s very relevant. So Andrei, that’s a really cool advice. I think setting the context right is more important than looking at metrics probably because that sets the fundamentals. But I would love to know how do you do that in your teams? Like, can you just like deep dive into it a little bit more so that we have some examples? Yeah?

Andrei Gavrila: Yeah. Sure. We are using something called maturity models. Maturity models got a bad reputation, some time ago, because they were used in a very heavy way, very, let’s say, non-flexible way. But we developed some maturity models that are really aligned with this agile mindset and are really one of the best tools that CTOs, engineering managers can use to answer those questions. Am I in the context where my team could perform or not? And these maturity models are very simple to understand. They contain items that you can say they are done or not. And the whole idea is that it will help you understand in which one of the four states you are on a certain dimension.

So, let’s pick security just as an example. You have a software development department, multiple teams, and you are asking yourselves, “Where are we on security?” So, with the help of our maturity models, with which are, as I said, very easy to use, you’d be able to say, if you are in a state of ‘starter’, that’s what we call it. And if you are in a state of ‘starter’, it means you are in a significant risk of failure. Or you could be in a state of ‘almost good’, which usually means that you are no longer in that immediate risk of failure, but you have a lot of waste in the process. And then, there is this column that’s called ‘good’, which means that you’ve passed that wasteful moment, and now you are actually working to optimize the system. The last one is called ‘very good’, in which if you are, you probably have competitive advantages because you do more than what your competition is doing. And you can think that we have these maturity models for everything, security, but we also have them for people engagement and now if you want, we can try to create such a very small maturity model between ourselves around people engagement. So, what is one thing that you think companies should do to no longer be in an immediate risk of failure around people engagement? What, what do you think companies could do to no longer be in an immediate risk of failure around people engagement? 

Kovid Batra: Like, they, first of all, like they can go back to the people and understand from them that what are those different areas where they feel uncomfortable, what are the things that are making them not perform or not be the best version of themselves. So we can go and understand from them. 

Andrei Gavrila: Exactly. That’s the kind of item that you would find in that column where we are not necessarily going to tell you how to do it. We’re not going to tell you, do it and ask them these questions and check these answers. But we would say, if you’re not asking periodically, asking your team how they feel, do they like what they do, what are their biggest problems, then most probably in terms of people and people engagement, you are an immediate risk of failure, meaning people might leave tomorrow because they are not happy with what they do. They don’t feel respected or this kind of stuff. So this is exactly how our maturity models are created and are populated by these kinds of items. 

Kovid Batra: So I think that’s, that’s perfect. You said there are four dimensions, right? So, one is security. One is people. Can you just name the other two? I think I got the context of what you’re saying. 

Andrei Gavrila: The dimensions are, are these columns like ‘starter’. 

Kovid Batra: Oh, okay. Got it. 

Andrei Gavrila: The second one is ‘almost good’, which is waste. The third one is ‘good’, which is efficiency. And the fourth one is ‘very good’, which is a competitive advantage. So, all the maturity models have these columns, but then we have maturity models on people engagement, on skills, on security, on data, on architecture, on infrastructure, on testing, on engineering practices and many more.

Kovid Batra: Yeah, right. I think I got your point. For every, like there are a lot of aspects of an engineering team. And for an engineering team to be successful, all these areas which we define, they should, like keep improving in these segments that you’re talking about. So.. 

Andrei Gavrila: Exactly. 

Kovid Batra: Perfect. I think that’s a very relevant thing. And I’m sure this comes with a lot of experience. 

Andrei Gavrila: Yeah. So, last thing that I would want to add here is that when people ask me, how are my teams performing, I feel that they are not efficient or they are not working out. First thing that I do is say, “Let’s try to fill this maturity model, and to understand where we are. And then once we fill them, I’m saying, see here and here and here.”

Kovid Batra: Right. 

Andrei Gavrila: You are way below, let’s say safety. So in a way, until you manage to fix this, you will have maybe just marginal improvements. 

Kovid Batra: Totally. Great talk. Great talk, Andrei. I think thanks a lot for sharing all these insights and relevant examples. I would love to have another session with you sometime because I think these few minutes are very short for us to learn all those things that you have done in your past. So, we are trying to do that 70% bit from the experience, right? So, cool. I think we’ll connect again for sure. Thanks a lot for taking out time today and talking to us.

Andrei Gavrila: Sure. Sure. It was really great being here. Thank you for the invitation. Have a great day. 

Kovid Batra: You too, Andrei. Thank you.