Mohammad Mizanur Rahman is a Director and the CTO of Brain Station 23, one of Bangladesh’s largest software and technology development services companies with a growing global business. Before joining Brain Station in 2013, Mizan ran his own company, which he termed jumping into entrepreneurship without knowing much about what the journey entailed. At Brain Station 23, he looks after the company’s technology operation along with several other operational aspects, a fascinating and complex role given that the company is not a typical tech product company. The role not only allows him to shape the operational strategy of the company, but also gives an insider view into the world of technology. With AI radically reshaping the world of tech development, this is a fascinating position to be in.
In this all ranging conversation, we talk about his path to what he is doing today, his work as the CTO of Brain Station 23, operational structures and strategies of Brain Station 23 as the company grows into nearly a 1000-people team, how Brain Station 23 is navigating emergence of AI and what he thinks about how AI will reshape the world of software outsourcing business, his management approach, lessons from his journey and much more.
Founded in 2006, Brain Station 23 has seen phenomenal growth over the last nearly two decades, growing from a tiny team of a few people in its early days to a giant of nearly a 1000-people company. The company has not only built a strong global business, it has also built a fascinating organizational model, figuring out how to scale without breaking organizational culture of agility and excellence, create leaders at scale, and constantly evolve with the market.
Growth is a challenge for most companies. The first challenge is breaking out of an apparent growth plateau that all companies inevitably encounter after a certain stage in their journey. The second challenge is designing an organization that can manage growth without breaking apart. An apparent shortage of high quality managers and operators make these challenges doubly painful for many Bangladeshi companies. Being in an industry that is in constant flux and has high competition for talent, this challenge is more pronounced for a company like Brain Station 23.
We look into how Brain Station 23 has innovated on organization design to address some of these challenges. To break out of the growth plateau, take risks, focus on people, and deliver consistently reliable services. To scale successfully, build autonomous small business units using a combination of meaningful autonomy, responsibility, and rewards. Create and empower leaders. Decentralize decision making with meaningful logic. And finally, evolve relentlessly.
We also talk in detail about what Mizan thinks about how AI will change the software outsourcing business, how Brain Station 23 is navigating AI adoption, and whether the company has an AI strategy going forward.
Building a consequential and enduring organization is a near impossible feat. One of my takeaways from this conversation is that Brain Station 23 has figured out the puzzle. To learn how Brain Station 23 has done it, you have to read the interview.
With that, I’ll leave you to read the full conversation. This is a fascinating read in its entirety.
Happy building! Enjoy.
Mohammad Ruhul Kader: Thank you for agreeing to this interview. Please tell us about yourself and your path to what you are doing today.
Mohammad Mizanur Rahman: Thank you for having me. Let me start with Brain Station 23's background and then go to my own background. Brain Station 23 started in 2006 with just two people. We currently have around 950 people. We have mostly grown organically, serving both global clients and local businesses, MNCs, and organizations. We work with startups that require rapid development and highly adaptable processes. We work with enterprise companies that require complex, process-oriented solutions. Over the years, we have built a process where we can deliver agile, startup-like services with enterprise-level reliability.
Regarding my own journey, I was born and raised in Dhaka. I went to Badda Alatunnesa Higher Secondary School and then Dhaka College for my HSC in 1998. After that, I wanted to study computer science and enrolled in North South University for my undergraduate studies in 2001.
Throughout my university years, I worked part-time for a Finnish company. Soon after graduating, I started my entrepreneurial journey. I didn't have much formal corporate experience in the traditional sense. Something that can be both a strength and a weakness when building a company.
Later, I felt that having institutional business knowledge would benefit my entrepreneurship journey. So I did an MBA, majoring in finance, from East West University. I always felt I needed to understand numbers, hence the finance major. This covers my academic education.
I also gained additional exposure through several programs. In 2013, I was selected as one of five participants from Bangladesh for a seven-week entrepreneurship exchange program supported by Oklahoma University and facilitated by the US government. During the program, I worked closely with the CEO of a US-based company and participated in various university learning programs.
During COVID, I participated in a two-month MIT boot camp innovation program. The program had participants from over 40 countries, and our team won globally.
From Brain Station, we also participated in the Stanford Seed Transformation Program as the only software company from Bangladesh. We were one of two companies selected from Bangladesh at that time, along with companies from India and Nepal. It was a year-long program. Part of the program was held at Infosys campus in India. Along with Stanford faculties, an advisor from Infosys also attended the program. This experience was very helpful for me and for our company's transformation.
More recently, since I performed well in the previous MIT program, I was invited to MIT for a one-week on-site entrepreneurship program.
These experiences have been extremely helpful for my personal growth, and I'm continuously trying to learn and improve.
Without a lot of formal experience, you're not always trying to fit everything into a pre-existing mold. You are not bound to any one way of doing things.
Ruhul: What are the upsides and downsides of starting a business without prior experience?
Mizan: Let me start with the downsides. Without formal experience, you may miss things that are standard and common knowledge. Things that experienced people already know. Without that experience, learning and figuring out the basics can take time. You may end up making rudimentary mistakes. What might take five minutes for someone experienced might take much longer for you. Not knowing can mean you need to invest time in learning things that don't require it, leading to a trial-and-error approach where some things fail and some succeed. Overall, you may need to spend more time learning.
The upside is having less baggage. This allows room for creativity. You can ask, "What if we do it this way instead?" When you're accustomed to a certain way of doing things, you might think this is how it's done. It can close your mind to experiment and exploration.
But there might be room for a more creative solution. The process you learned might not be right for your current situation. When we think we know something, it becomes hard to unlearn it. When you start with experience, you risk operating from a closed mindset where you don't entertain alternative ideas.
You try to fit everything into what you learned. You have a few frameworks that you try to superimpose everywhere. On the contrary, without a lot of formal experience, you're not always trying to fit everything into a pre-existing mold. You are not bound to any one way of doing things. You are more open.
Ruhul: This is a wonderful point. The curse of knowledge. Knowing something can become a limitation. You may feel that I know this, and this is how it must be done; I won't consider other ways. That's often a limitation of knowledge. Why did you decide to become an entrepreneur? What was the motivation? When you started your first company, these were the very early days of software companies in Bangladesh; it was not especially easy to build businesses in Bangladesh.
Mizan: Raisul and I have known each other since 1999, for almost 26 years. We are friends. To be clear, I wasn't there when Brain Station started. I had my own company. I joined Brain Station in 2013.
Regarding my path to entrepreneurship, as I mentioned earlier, I worked part-time for a Finnish company throughout my university studies. North South was expensive. I feel proud that I eventually paid my tuition fees with my own salary from that part-time job. That was quite significant for me.
My path to entrepreneurship wasn't a conscious decision; it was not like I had to start my own company. It happened organically. I always loved trying interesting things. Everyone else seemed to be chasing jobs, but those things didn't attract me much. It was more like let's do something interesting and keep doing it, and if we do it right, it will grow. I didn't think much about the risks or anything, to be honest. I started it as a small experiment. It started doing well, and I started feeling good about it. I figured, why not do it some more? That's how it was. I don't see it as some big thing to talk about. It just happened naturally.
An important aspect of work for me is whether I'm enjoying what I'm doing. Part of that enjoyment comes from the people you work with.
Ruhul: You started this company, ran it until 2013, and then decided to join Brain Station. Tell us more about that journey.
Mizan: Working at that Finland-based company during university gave me exposure to the industry and the world of software. It was an excellent learning experience. Moreover, I had friends who worked with me in that company. We would finish classes, go to the office, work till 8 or 9 at night, then play games all night, and then go back to class in the morning. It was quite undisciplined. We did whatever we felt like. But it was great fun.
After graduation, I started my own company, mostly serving foreign companies. It started differently. Initially, I worked on a platform called GetAFreelancer (which later became freelancer.com). I started by doing small projects. This exposure was very interesting for me because clients appreciated the work, and I was earning. My rating was good, around 4.93 out of 5, which helped with new projects. That expanded into work outside the platform. I did some work on my own. I built a small team, and that's how it went.
Raisul and I used to meet regularly, talking about our work, various issues, and challenges throughout this period. Entrepreneurship is a lonely journey. Sometimes you feel you need someone to share your joys and sorrows. Someone who can relate to your journey. It's difficult even for a best friend in a traditional job to relate to the pain you go through as a founder. The helplessness, excitement, inspiration, and constant emotional ups and downs. It's hard to connect when you don't go through the journey. So we met and shared our work.
In these meetings, Raisul used to say, "Let's work together and build something amazing together." That discussion eventually led to my joining Brain Station in 2013. When I joined, we had a team of around 40-60 people. Alhamdulillah, we have grown from there to a team of over 900 people.
Ruhul: I remember having these conversations with Raisul bhai. And he would always mention how important it is to work together if you want to build something meaningful.
Mizan: Not because we are speaking about him or because he is my friend or partner—I say this from a place of humility, Raisul is truly amazing. An important aspect of work for me is whether I'm enjoying what I'm doing. Part of that enjoyment comes from the people you work with. When you work with a friend, you can ask for help and suggestions. Share and discuss challenges and difficult times, as difficult times will come no matter how good you are.
When you continuously strive for excellence, being able to collaborate with someone you understand and have an excellent relationship with can be invaluable. So being able to work with Raisul is a privilege.
The more interesting thing is that our CFO, Raj, is also a close friend. We come from the same batch, which makes working together more fun. Work never feels like work.
The growth we have achieved is a result of this team effort. Moreover, Allah has been very kind. Many times, wrong decisions were made, or we took major risks, or didn't put much effort, but somehow things turned out well.
Ruhul: Working together with your friends, especially with ambitious people of the same caliber, can be challenging as well. Maintaining relationship dynamics can be challenging both during good times and when things become difficult. How do you deal with that?
Mizan: That's a good point. I've seen many times that even close friends can encounter challenges when business is involved. Maintaining a relationship, especially during difficult times, can be challenging.
In our case, a major factor that worked is mutual respect. This means humility, being a good listener, and the capacity to accommodate the point of views and limitations of others. Everything will not be according to my wishes, but that's alright. Raisul is very good at these things. He has played a role in establishing these values.
When you are working together, you have to accept the fact that everything will not be 100% right. In many instances, you don't need to get involved with everything. Being able to let things go is important. Nature will autocorrect. Many times, we can't let go. We try to micromanage that leads to problems.
But when you learn to let go, those areas of conflict significantly reduce.
Be more generous when the other person makes a mistake. We have to understand that mistakes will happen. Instead of focusing too much on others, we should self-reflect on our own limitations. Mistakes will happen, and that's alright. Different people will contribute differently. These attitudes have played a major role.
We have also put systems and processes in place to take pressure off our personal relationship. Decisions are made mutually. In areas of high risk, we discuss together to assess potential impacts on organizational sustainability. It's often difficult for people to accept when the discussion turns to "he made this mistake, so this happened." When we blame each other instead of taking a neutral approach, that's when it becomes a challenge. We have created systems to overcome these tendencies.
Be more generous when the other person makes a mistake. We have to understand that mistakes will happen. Instead of focusing too much on others, we should self-reflect on our own limitations.
Ruhul: This is beautifully put. Being able to accommodate each other. If you want to maintain a good relationship, be it in business or in personal life, you have to overlook the other person's shortcomings. Alain de Botton calls this charity of interpretation. Being emotionally charitable. You interpret others' shortcomings flexibly, which reduces the weight of their mistakes and makes it much easier to handle. As you mentioned, Brain Station really started growing after 2012. In your observation, what has contributed to this growth?
Mizan: One major factor is the intent behind starting this company. Many companies focus on revenue or profit, but we've always wanted to employ more people. Can we become a 20,000-person company? Not in terms of revenue, but in terms of people—can we have 20,000, or 30,000, or 50,000 people so that we can impact more lives? Our goal has always been to impact the lives of more people and their families. This orientation has guided many of our decisions as a company.
When HR tells me they've found two good, albeit expensive, developers, we almost always try to bring them on even if there is a project at the moment. We think we might not have a project right now, but it's good to have strong resources so you can serve when a project comes. We hire freshers twice a year.
These are major investments that impact our profitability. Many companies prefer to maintain a high percentage of billable resources to maximize revenue and profitability. For us, as long as sustainability is not at risk, we prioritize becoming a bigger organization with more people over profitability.
I wouldn't say whether this approach is right or wrong. But it has helped. It gives us flexibility to take on opportunities when they come our way. When a client asks for a certain resource, we often have free people available, and we can immediately take on the project.
Companies that are very tight, only onboarding new people when a billable resource is needed, often miss out on projects that come serendipitously. You can’t plan everything. Many clients come with projects for which they can’t wait. If you say it will take one or two months to put together a team, it doesn't work. One upside of our strategy is that we could take on opportunities as they come. We have more flexibility to expand. Opportunities will always find you when you are ready.
For instance, it happened that a client with a 10-person team discontinued, but then an opportunity came for a 45-person team. If the client who needed 10 people hadn't discontinued, we might not have been able to take the new opportunity.
As I've said repeatedly, God has been so kind; in many instances, things happened for us. We have a saying in the team that a client discontinuing means something new and bigger is coming. Our thought process has changed. When you believe something strongly, it happens.
This played a major role: continuously increasing the number of people.
We never hesitated to add people to the team when we could do it. Adding new people is costly and risky. Sometimes we hired people without a strategy. A logical person might question our decision-making process and would be very upset. They would say that you are not making thoughtful decisions. You are not looking at company sustainability. In the beginning, in many of these instances, we didn't really think about sustainability; it was more a certain kind of faith that things would work out, God willing.
Our goal has always been to impact the lives of more people and their families. This orientation has guided many of our decisions as a company.
Ruhul: This is interesting. Faith, of course, plays a role. The world functions in a mysterious way. But it is also a function of courage. Faith is an expression of courage. There is a real risk when you are expanding your team.
Mizan: We're more afraid now to take risks because with so many people, a wrong decision has a significant impact. It was easier when we were smaller. The more you grow, the more logical you become. You become less of a risk-taker. That's alright as well because when you have bigger responsibility, you should be more judicious in taking risks.
Ruhul: This is an interesting perspective. By extension of that, please talk about how organizational decision-making changes as a company grows? And what are the challenges of managing the growth?
Mizan: Looking back, when we were 60, I knew everybody, even their families. They knew me. I could say who was working on what and their project statuses on the go. It was easier to track people and things. As we grew, things became increasingly complex.
When we were smaller, HR decisions were easier. You can easily assign people something and follow up. As you grow, that's no longer feasible. Getting complete visibility of things gets increasingly challenging. You need to put together a system, KPIs, tracking, rules, guiding principles, etc. When we reached 100, we needed to establish some processes.
Now, with 950 people, there isn't enough visibility. It's hard to know if someone assigned to a task is doing it. It is hard to know whether guiding principles are not being followed properly. Proper appraisal becomes a challenge. People who are good at selling themselves can get away with undue benefits just by talking more about their work, even if they are not the best performers. Whereas an introvert doing excellent work might not be recognized or appraised properly, get frustrated, and could suddenly decide to leave. Then you realize what you have missed.
As the company grew, I didn't really know people one layer beyond my immediate circle. Many people don't even know me. We need to be mindful of these challenges as the company grows.
The key is to make things work without first-hand interaction. We had to build these layers to ensure communication flows freely and effectively from top to bottom. Not having experience in a formal big organization, we had to learn by doing. We learned by experimenting with different processes, seeking expert guidance, and making a lot of mistakes.
At every step—from 60 to 100, 100 to 200, 200 to 400—the previous way of doing things stopped working. Continuous evolution was necessary.
When we were a small team, the culture was strong. We knew these people loved the company. As the company grew, we realized that with a larger team, we needed to ensure everyone understood company values. We had to find an effective way to communicate these values.
This led to the emergence of business units with their own program managers and project managers. We had to think again about how to do appraisals, how to communicate, and how to make the operation smooth.
As an organization grows, you have to accept many things. Some people might not be fully aligned with the organizational value system. You have to think about how to address that, but tolerate the divergence. Many things would not happen as you expect.
When these challenges occur, we try to understand and solve them systematically. We focus on the support system rather than blaming individuals. If a mistake happens, we accept it the first time. We try to learn from it. Then we try to make an air-tight system to ensure it doesn't recur.
When you are a small company and a mistake happens, you can be specific about who made the mistake, and you can work with that person. When you grow large, it has to be about the system. A bug in the system is causing this problem because it couldn't prevent it from happening. You see things from that perspective.
We adopted the EOS (Entrepreneurial Operating System) process to facilitate more systematic collaboration. Every week, our leaders have two 1.5-hour meetings. One with their next-level leaders and one among themselves. They, in turn, meet with the layer below them so that information gets transmitted. You maintain a consistent organizational flow.
This process worked perfectly up to 500 people, but for larger scales, we needed a scaling process. We have now adopted the scaling-up process.
In our weekly L10 meetings, team members communicate, share numbers, and discuss issues using a framework called IDS (identify, discuss, and find solutions).
During COVID, we closed the office on day one. We invited all leaders to share their thoughts and then collectively decided. This is a decision-making framework that we use. Everyone gets involved in the process. Once a decision is made, everyone takes ownership regardless of the outcome.
There's no silver bullet. Challenges are constant. As the company grows, dynamics change. New challenges come. Our work is also constant, trying to find solutions to problems.
I see many large companies going through a similar trajectory. Many of these organizations have mature systems. Mature systems, while useful, also have some drawbacks. Bureaucracy creeps in and slows things down. Lack of clear authority and the rigidity of the system make doing things difficult.
These don't align with our organization's value system. We prefer speed and zero bureaucracy. Still, some processes are required so that things run smoothly and you don't need to do the same thing twice. However, the purpose of these processes is to speed things up, not to slow them down. Our processes are designed with that in mind.
One of our core values is speed—work fast. We work like a small company so the negatives of big companies don't affect us. That's why we have divided the organization into small business units with independent leadership teams so they can operate like small organizations and respond fast.
The key is to make things work without first-hand interaction. We had to build these layers to ensure communication flows freely and effectively from top to bottom. Not having experience in a formal big organization, we had to learn by doing. We learned by experimenting with different processes, seeking expert guidance, and making a lot of mistakes.
Ruhul: This is very interesting. Can you expand how you maintain agility as a large company? How have you created a mechanism like SBUs (Small Business Units) where they operate independently? How did you set these up? What was the thought process behind this model, and how does it work in practice?
Mizan: We came to see that as the organization grows bigger and everything centralizes, things start failing. But this centralization is not always necessary. CXOs don't need to deal with every decision. Everything doesn't have to come to one point. This centralization not only slows down the organization, but it rarely helps you make the best decision. If the same person handles e-commerce, fintech, pharma, and telecom, they can never make the best decisions in all these areas. They wouldn't be able to think about the problem in a focused manner.
The second aspect is that many companies focus on a single niche. People usually say that when you do too much, you run the risk of spreading yourself too thin. That niche is a better strategy. It allows speed and depth.
We figured we'd try to find a balance between these aspects. We will pursue niches, and we will also pursue combinations. That's how our SBU model came into being.
We thought, what if there were small business units that could operate with relative independence except for a few aspects. We didn't overthink it and followed a simple approach. We found people working on different verticals and converted them into specific business units. With people working in Telco, we created a unit for Telco. We appointed someone from the team with leadership potential to run the unit. Similarly, we created a few more similar units, such as a unit for ecommerce and one for European markets. We thought, let's experiment and see what happens.
So we have these different SBUs. Some are functional and domain-specific, such as ecommerce or fintech, that deal with those specific verticals. Then there are SBUs like Europe that deal with geographic markets. But that's alright. The main thing is creating these smaller units. After dividing, we logically separated their financials and other aspects and gave them logical autonomy.
How does this work? Each SBU has completely separate finances, with full visibility into its expenses, earnings, and profitability. The SBU head has full visibility into the performance of their SBU. They run it like a business.
The main beauty of this is that when I talk to the e-commerce SBU or the Telco SBU, they discuss innovations in their specific domains and suggest expansions or additions.
Since the SBU head is empowered, they make decisions. If a discussion is needed, we discuss, but there are no restrictions. They have all the data—organizational, resource, and unit-specific. They have as much information as I do. For an SBU head, their 80-person or 40-person team is their company. They know everyone.
You may consider these as small companies. In the future, we may create subgroups within these units, like an ERP team within the e-commerce group, which would have its own leader. Many groups will be formed, each working like a small company.
What used to happen was that for small things, they would come to me. Now, even if I'm away for a week, things won't stop; they'll most likely run better. If a discussion is needed, we say "discuss," not "decide." The decision lies with them. They can discuss it if needed. Now we have a lot of smaller organizations under a large organization that work in a focused manner, working on a specific niche.
As an organization grows, you have to accept many things. Some people might not be fully aligned with the organizational value system. You have to think about how to address that, but tolerate the divergence. Many things would not happen as you expect.
Ruhul: This is beautifully done. Large organizations face many different challenges. This is more beautiful because you've separated the finances. It means you have a clear picture of whether an SBU is profitable or not, whether we should continue it or not, etc. The SBU head can decide what they should do.
Mizan: SBU heads actually do the forecasts. Each SBU head completely decides how many resources they will need the next year, and the finance team continuously provides tracking data on planned value, reality, and variance. They get this data weekly and monthly.
Ruhul: I think this itself can be an hour-long discussion. But let's get to your job as the company's CTO. Brain Station is a very different company from a typical tech product company. Accordingly, I think your role as the CTO is very different from traditional product-led companies. Tell us about your work as CTO at Brain Station. What does a typical day look like for you? How do you divide your time between strategy, operations, and innovation?
Mizan: This is interesting. Normally, the roles a CTO has in a startup organization and my role as the CTO of Brain Station don't have many similarities.
As we've seen, every business unit has architects and senior technology leaders. Technical challenges usually don't come to me. When they do, they are mainly in the area of strategy, and when things escalate. Alhamdulillah, we rarely deal with escalations that demand my involvement.
More regular involvement is in strategy. Regardless, how should we approach HR and so on? We have created a group called the Center of Excellence (CoE) to discuss many of these things. A major aspect is how we enable our people so that first-hand contribution and technology involvement don't need my direct input. People are already empowered. If training or certification is required, they get it.
From a CTO perspective, I don't spend 80% of my time on technology, which might be common in other instances. Rather, many different things continuously come up.
As a certified partner for LMS Moodle, they might have upcoming events and want me to join along with CTOs of other partner companies. It's more about collaboration and working with other organizations when it comes to organization-building-related work.
It's also about having someone for technology-related issues. If something doesn't work, they have someone to go to and ask for help. It doesn't mean I try to solve it myself. I don't. Being the CTO, I have visibility of who can better solve a particular problem, so I involve that person.
I never try to solve a problem myself because if I do that, it can become a bottleneck. If I try to solve every problem, I become the bottleneck. This is a major change. We have tried to bring this approach to our next layer. We tell people to involve their next layer in the problem-solving process so that knowledge gets transferred, and your next layer has the exposure you are having. Otherwise, the difference between layers becomes so high that, despite similar capabilities, people can't grasp the context due to exposure differences.
Being CTO, you get to see many things that your next layer can't see. The problem with this setting is that, as a result, they can't contribute effectively to solving problems. Perhaps you are dealing with a particular challenge. Since your next layer doesn't have exposure to it, they end up suggesting solutions that you may find irrelevant.
My main focus is to ensure that whatever knowledge and learning exists flows freely and is distributed to different units within the organization's technology capacity.
I take the initiative to share when a unit does something interesting so others can learn. Since I can see what's happening across different units, I can pull things together and create an environment where people share their ideas and knowledge across the organization.
On top of that, with AI and other disruptions, the organization needs to figure out support systems to help our people navigate the new environment. We might need infrastructure to run models, so it's about collaborating on these things.
Much of my input is more on the strategic side and collaboration. Since I've had an interest in technology from the beginning, I share my learnings if it helps. It's not like the traditional CTO who does the architecture and does the product. We don't need to deal with those things. We want the knowledge to be there so that when I communicate it, there's clarity and things don't change in translation.
We came to see that as the organization grows bigger and everything centralizes, things start failing. But this centralization is not always necessary. CXOs don't need to deal with every decision. Everything doesn't have to come to one point. This centralization not only slows down the organization, but it rarely helps you make the best decision. If the same person handles e-commerce, fintech, pharma, and telecom, they can never make the best decisions in all these areas. They wouldn't be able to think about the problem in a focused manner.
Ruhul: Your technology function is relatively different from what we might call a traditional tech company. We can understand a bit of that, as you mentioned that each SBU has separate capabilities. For example, when we talked about cloud yesterday, you said every team has cloud expertise.
Mizan: Yes, yes.
Ruhul: So, if they need cloud capabilities, the SBU in fintech, which deals with fintech, also has cloud capabilities. This is one aspect. And then there's the technology function, the structuring of the services you provide to other companies. And as you said, there's a learning aspect—strategically thinking about where I might need to invest in learning. Philosophically, it also comes down to how you build and lead the team. It is a pretty complex scenario. You partly answered this, as you mentioned, it comes down to focusing on creating the next layer of leadership so that when a problem comes, they can handle it. This is a philosophical approach where you are constantly building a pipeline of leaders who will take over, so that the company can scale effectively. Make sure I don't become a bottleneck instead of a solution. You do all these so you can deliver great services to your customers. You also have your technology function, where you have tools, systems, and processes. Perhaps you are not building a product yourself, but you are building hundreds of products for other people. Your people need tools and tech to do these things. If you think about the entire technology function of Brain Station, and then think from a philosophical perspective, what would that philosophy look like? I hope I make sense.
Mizan: We've recently created a group called the Center of Excellence (CoE). In the CoE, we have about 90% technology managers, managers of managers, who already oversee the technology side of multiple projects, along with some project managers. We recently had a two-day session outside Dhaka with CoE and business units to understand that while SBU heads will focus more on strategy and collaboration, the CoE members will look into operational and delivery excellence.
One major focus is fast delivery of strong technology—how fast we can deliver technology solutions and how good we are at doing it.
The interesting thing is that all our SBU heads are from a technology background, which is an advantage. They are not just management people. They understand the vibe of engineers. We have 20-22 CoE members from different SBUs, appointed as leaders, so they can look into the technology excellence and thought leadership for multiple teams.
Our CoE has a weekly call where members share new tools and methods that they use to solve problems. Someone shared an experiment where a certain thing that took five days now takes two days or four hours. This encourages others to try it. If a subscription or tool is needed, the SBU head has full authority and capacity to buy it within a certain budget. In many instances, people are taking subscriptions, but I don't know about it. They have separate cards for purchases. Since they have full visibility of finance and profitability, they automatically consider the impact of expenses. They might decide not to give everyone a subscription if it heavily impacts profitability, and instead, find a smaller-scale solution.
The SBU head, along with CoE members, makes these decisions. They ensure spending decisions are made logically. These are not management decisions. We completely rely on these people to make the right decisions.
The CoE fills a gap where strategic business leaders often had to oversee too many strategies and operations, making operational excellence difficult. Before CoE, this was a challenge for SBU heads. The CoE acts as a supporting hand for continuous innovation and new ways of doing things.
We plan to grow the CoE from 20 members to 50-60 people, creating a proper pyramid of leaders. We now have over 120 leaders across the organization.
It's also a good question: how do you create leaders and how many? For us, each SBU has quarterly and yearly KPI targets for creating new leaders. They identify how many leaders they have and their targets at the end of a certain time. This comes from the understanding that if leaders are not developed, we can't scale. Capable supporting hands are essential.
At the same time, helping existing leaders grow through continuous learning. We continuously conduct various training sessions to build this knowledge. Otherwise, you wouldn't be able to scale. Are we fully happy with it? Not yet. It's not easy either. The direction is clear, but execution is the challenge, ensuring even the newest joiner gets the impact.
We have guiding principles, and when discussing different technologies, approaches, and policies, we check if they are aligned with these principles. If there's an exception, we discuss it separately. We don't expect everything to fit our principles. But having these principles is helpful. You can make regular decisions faster. So, guiding principles are there. The value system is there. When something new happens, as long as it is not a risk for the organization, we allow decentralized decision-making.
I never try to solve a problem myself because if I do that, it can become a bottleneck. If I try to solve every problem, I become the bottleneck. This is a major change. We have tried to bring this approach to our next layer. We tell people to involve their next layer in the problem-solving process so that knowledge gets transferred, and your next layer has the exposure you are having.
Ruhul: We talked about AI several times. Yesterday, we talked a bit about how the outsourcing industry will change under the influence of AI. Since you are in the middle of it, how do you think AI will change and reshape the outsourcing industry? What are the threats and opportunities you see because of AI?
Mizan: With AI, many things are now more possible. Someone was saying that now even one or two-person companies can become billion-dollar companies. You will have a proliferation of many more initiatives, digital initiatives.
Before, a fresher's quality baseline was quite random. Now, the baseline, the benchmark, will be much higher automatically. People won't get away with doing low-standard work. They'll need a minimum level of quality, capacity, and speed.
The way of doing things is changing. Let me give you an example. Creating SRS (Software Requirements Specification) and RFQ (Request for Quotation) responses used to take one and a half to two weeks for large projects before. With a tool called BMAT, it is possible to do it in two days, which includes manual human review. The initial AI-generated version takes about three hours. Then, human reviewers check it, sometimes consulting another AI, creating a validation process.
What I primarily see is that we have to accept that things won't remain as they used to be. All the steps of the Software Development Life Cycle (SDLC) are changing. People's expectations are also changing. Before, a registration or login page taking one or two days was acceptable. Now, customers might reject such timelines.
Resource augmentation used to be a major area for us, with people working as team members with different companies. We are seeing that customers are becoming more solution-driven. They are looking for solutions instead of resource augmentation. New features can be added or dropped. They may eventually hire a team for maintenance that can work on a monthly resource augmentation model. The total flow has changed.
Customers expect us to come first with something that will give a kick start on day one, which will also validate your knowledge and understanding of the problem domain. These areas are changing rapidly.
Before, value addition was separate from coding. Now, even developers need to know and understand the impact – whether it's good, whether it is the right way of doing things, why it's being done this way, and its potential impact.
Engineers often don't want to think about impact. However, as I see it, with time, execution will be less about manual coding and more about deciding how to execute using tools. A high-level technical overview will be more important. The development timeline will also be compressed.
However, this doesn't mean AI will replace programmers. It will help. Technical background will remain critical to instruct and think in multiple layers. Without a proper technology background, you wouldn't be able to do it right. What will happen is that greater skills will be needed. Programmers will also have to understand business relevance.
A lot of disruption will be happening. AI will make many things more feasible. A single person building a multi-billion-dollar company would be very possible. Many big companies like TCS or Accenture might see a group of people becoming irrelevant.
Part of the reason is that many large companies usually have a high level of specialization. People might become specialized in only one task and do nothing else. That was alright. It improved efficiency. But with AI, it wouldn't be alright anymore. Many of these people might not be reusable as AI can do it better due to its strong knowledge base.
When you are a large company, you probably have 50,000 people in customer support. Large companies need these scales to set up. But with AI, some of these tasks can be done differently, and you can find it difficult to accommodate so many people. You need to think about replacing some of them to maintain efficiency. This scenario is probably coming. The repetitive tasks and jobs would be heavily tested.
We have guiding principles, and when discussing different technologies, approaches, and policies, we check if they are aligned with these principles. If there's an exception, we discuss it separately. We don't expect everything to fit our principles. But having these principles is helpful. You can make regular decisions faster. So, guiding principles are there. The value system is there. When something new happens, as long as it is not a risk for the organization, we allow decentralized decision-making.
Ruhul: This is fascinating. One sense I get is that the way we operate and do things is going to change. Our tasks may also change. We would probably be doing something entirely different than what we used to do. We will probably move up to doing something more complex and higher value. My next question is whether these changes will also change the business model of the outsourcing business?
Mizan: Yes, it is going to change significantly. But there are nuances here. You would get an email from time to time that your brain would automatically detect as written by ChatGPT. Then it fails big time.
I was having a conversation with one of our clients who was complaining that he had asked one of their vendors to provide a solution to a problem, and they sent him a long answer that didn't say anything decisive. He was so pissed that he told them that most likely, you are not taking me seriously. Because he was asking for an answer, but they sent a beautifully written response that didn't mean anything.
The point is that tools will not be able to solve all your problems. You need the capacity to filter and curate things in a manner that serves the purpose.
I need to give the right prompts and instructions and have the knowledge to do so. A person with domain knowledge will give instructions from that context. A good content writer understands what the readers want. Without that knowledge, AI might produce something beautiful but lacking value.
Someone who understands this can use these tools easier and faster, with significant benefits. What used to take them three or four days to think about, they can now deliver in two hours because they know what to do and can use the tool accordingly.
A skilled person who could previously produce 8 out of 10 might now produce 12 out of 10 because it's more possible. They can try many different approaches and try them fast. They can perform A/B tests easily.
What used to be expensive thinking this way, they can now create multiple sets, test them, and use the data for effective large-scale operations. Execution time is greatly reduced. As long as we know what we want to achieve, execution can be much faster.
However, it is also important to understand that it's not about things becoming fast; it is that the whole thing will change. It's not only like something that took us five hours before, now we could do it in one hour. We might address the problem in a completely different way.
The SDLC components themselves may change. It's a different level of change, like from a horse to a car. We don't yet know how things will change at the end of the day, but the more I observe, the more I see that we are entering uncharted territory which is both exciting and frightening.
Ruhul: Have you already seen any changes in how you operate the company, how you deliver services, and your clients and client requirements? As you said, all kinds of things are being said about the future of AI. Many say AI is taking over programming, and many software jobs will be lost. What do you see happening in reality in this regard? In reality, if this happens, Brain Station will become a completely different company. Maybe a relatively less number of people company, but maybe much more valuable.
Mizan: You've raised an important point. In our next ten-year target, the number of people we have in the team is an important attribute. We have always wanted to have a large team. However, we are now thinking whether we should revisit our strategy and targets.
Regarding your question, we're seeing that no one wants to miss this AI trend. A good portion doesn't know how to leverage it – where, why, how. Many are slowly adapting. Consulting companies are starting to play a good role, analyzing businesses and suggesting areas for AI implementation. These changes are leading to incremental improvements. Manual processes are now done with AI. Efficiency is changing. Customer experience is changing.
This is a natural change where companies are integrating and leveraging AI in their work without making any other major changes. This is also happening in our operation. Teams are leveraging AI across how they work. They are also collaborating with clients to help them integrate AI into their organizations.
Secondly, customer expectations are also slowly changing. Before, a customer might be open to something taking 10 days. Now, they ask, "Are we really leveraging AI enough if it takes 10 days?" Others are more specific—can we do this with AI so that execution time and cost can be reduced? This change is already happening.
There are also areas where we are executing things faster. The example I mentioned earlier, where we produced a document in two days, which used to take much longer before. The first version of a document was produced using an LLM (Large Language Model) in three hours. Now, some customers might think it can be done in three hours, but that doesn't include human review. It depends on the customer's maturity. The document can be prepared in three hours. But if it's not reviewed, there are chances of mistakes, and there are future consequences if that happens. There are issues like this.
So, both sides need education. The speed of change is so fast that even we miss development from time to time. When a customer asks whether we have tried this or that tool, a tool that could speed up our work, we realize we have no idea about it.
The one strategy that will work for now is being more open to new things. Even a non-technical person's suggestion should be appreciated positively. This was useful before, but not necessary. But it is critical in today's context. We must be open to learning new things. If you are not open to learning from all kinds of people, you risk falling behind.
So, a learning mindset is needed on both sides. Expectations are changing. And enterprise organizations are cautious, leading to slower change. Startups and smaller companies are more aggressive in adopting these changes, focusing on fast execution. However, the speed isn't always the highest because if relevant stakeholders aren't ready, it doesn't matter.
I visited a startup competition where I came across a company building mobile apps for farmers. I asked who the end users were. They said mostly farmers. I asked how many farmers have smartphones. They said only 3-7%. The technology itself was bottlenecking their speed. Sometimes, the question is: Is my target customer ready for what I'm doing? Even if I develop an application in a day, it doesn't matter much in reality if the target customers aren't ready.
Regarding AI, changes are happening. Change is inevitable and will continue. Mindsets are slowly changing, but technology's speed is quite high, and adaptation isn't able to catch up with the speed. Tech is moving fast, but for transformational change, adoption needs to catch up, and it seems it will take some time. But it will happen.
The SDLC components themselves may change. It's a different level of change, like from a horse to a car. We don't yet know how things will change at the end of the day, but the more I observe, the more I see that we are entering uncharted territory which is both exciting and frightening.
Ruhul: This is a fascinating perspective that one of the ways you can navigate this change is by being open to learn from everyone. People you perceive as someone who is not technically literate can also teach you something. This is an excellent philosophy when you are in the middle of a massive change and in need of constant learning.
I'm asking a repeat question. Do you see any significant change in your operation in terms of productivity gain because of AI? Without AI, you might have to hire 10 more people for a project, which is not required now because you have AI. Is this happening already? Do you see any significant productivity gain where you are able to solve complex problems faster?
Mizan: This is already happening. Our different teams are adopting it, and it is more efficient. We run some internal surveys from time to time, and they show 25%, 30%, and 40% efficiency gains from using AI tools. Some of these findings are somewhat vague. What used to take four days can now be done in three hours.
The problem is that some of these problems themselves are new in nature, which makes it hard to accurately gauge whether AI adoption is the main reason behind productivity gains. However, there is a general sense that things are moving faster and there are real efficiency gains.
However, this is the current situation. In the future, when AI models are much better, I'm concerned about how much of an edge humans will have. The standard of human output will need to be much higher to compete with AI. AI will eat away many more low-hanging fruits. There is potential risk going forward. Many positions will cease to exist, but new positions will be created.
Increasing maturity means people will be more efficient. IT might not be the guaranteed job provider it once was. But people leveraging technology for farming or manufacturing might see increased demand. Things will keep shifting, and continuous adaptation will be the key. If we stubbornly stick to what we are doing and resist the change, then there is a chance to eventually become irrelevant.
Regarding AI, changes are happening. Change is inevitable and will continue. Mindsets are slowly changing, but technology's speed is quite high, and adaptation isn't able to catch up with the speed. Tech is moving fast, but for transformational change, adoption needs to catch up, and it seems it will take some time. But it will happen.
Ruhul: What is your organizational AI strategy? Do you have a long-term AI strategy?
Mizan: We are continuously trying to leverage AI and enable it across our operations—HR, finance, operations, etc.
We are constantly communicating the importance to the entire organization. There is not yet a single way of doing it that you can mandate. The whole situation is quite fluid. What is needed is for people to understand that they can do their current work better and faster by leveraging AI.
The Center of Excellence (CoE) that we have created is part of an effort to create an organizational culture where we are constantly reinforcing this culture of seeking new ways of doing things and embracing change. We are continuously communicating this message. We believe that is the only meaningful way to execute organization-wide change.
Things are changing so fast that templates become irrelevant. Instead, we convey the intent: be open, continuously look for ways to accommodate new changes. This has led to teams finding specific, beneficial solutions that we, from a higher level, might not have thought of.
Different Strategic Business Units (SBUs) are taking sessions, sharing case studies, and sharing good findings in leadership meetings. Communication is the most important part. You can't make everyone adopt certain changes. Those who embrace it will adapt, others will not. Some people will refuse change easily, and that's fine. They will eventually embrace change, or they will become irrelevant, or try to find a different field of work.
The Center of Excellence (CoE) that we have created is part of an effort to create an organizational culture where we are constantly reinforcing this culture of seeking new ways of doing things and embracing change. We are continuously communicating this message. We believe that is the only meaningful way to execute organization-wide change.
Ruhul: That's a beautiful strategy because this is such a fluid situation, you can't address it with one template. If you stick to one way of doing things and don't change, you will eventually fail. You need to have a mindset where you are constantly learning, upgrading yourself, and adapting to changes happening every day.
Your position is interesting as the company's CTO. On the technology end, you have a closer view of things happening in the tech world. At the same time, you have been building the company and have an exposure to the operational aspects of the company. If you look back on your journey, what are some of the biggest lessons in terms of organization building, technology, work, and life?
Mizan: The first lesson is we all make mistakes. Making mistakes is alright. The important thing is whether we learn from those mistakes. When we put our mistakes to better use by learning from them, that is the best way to learn anything meaningful. When we allow ourselves and others to make mistakes, we empower ourselves to get things done.
It is more important to get things done than to seek perfection. Imperfect execution is far better than perfect nothing. Striving for perfection can be a serious blocker. Many times, we spend so much time seeking perfection that it becomes a bottleneck. Getting things done is important, and doing them on time and within budget is important. Once you get something done, it can then be improved upon. We must pursue high standards and do things as best as we can, but it must not become a barrier to getting things done.
The second lesson is empathy, which can reduce many misunderstandings. Empathy, along with curiosity, can be transformational in our personal setting. Sometimes we get annoyed without being curious about what actually happened. Someone missed a meeting and our general tendency is to get angry, but it could happen that he faced some challenges that we don't yet know. Instead of trying to understand, we come to a conclusion, or it could be that he found a better solution to the problem. When we remain fixated on our view of the world without trying to understand the perspective of others, it can create problems. But the opposite is much better if you are truly seeking a solution.
As humans, we prioritize our own problems. We think that we are facing the biggest problem in the world. But I've seen many people working here with far greater personal challenges, like taking a sick family member for dialysis three days a week. These realizations help us not be judgmental.
Practicing empathy or being curious in a moment of personal disappointment is not easy. But as I have tried to practice these values, I am getting better at them. Now I try to catch myself when I see I am falling into the trap of being judgmental or being too critical, and ask myself to review it. When we practice these values regularly, these muscles become stronger, and it allows us to see the world in a fuller view. Now, when I get angry, I also try to see the reality.
Another lesson is being honest. Sometimes it's difficult to say things when you are honest, but compromising on honesty loses trust. Moreover, when we are not honest, we are constantly in need of making adjustments based on the situation.
Always maneuvering, we lose connection with our true selves. It is a problem for the self. When we are not honest, we try to sugarcoat a challenge and, in the process, lose the intent of the whole thing that eventually leads to an even bigger challenge.
I have seen it is more important to be genuine and authentic. It might be uncomfortable at times, but it is always a better strategy in the long run. Whatever corner we cut in life, they eventually come back to bite us.
With my leadership team and colleagues, when I screw something up, I'm always ready to acknowledge that I failed. I accept my failure and promise to try better next time. This strength to accept your failure without fuss is something that I had to learn, and it was not easy initially. But as I started practicing it, I came to see that it is a much better strategy. When you are honest and you share your limitations without hiding them, other people in your team can help you better. If I am not truthful, the problem will not go away, and no one will be able to help. But when I say, I am struggling, can you help, then it opens new doors.
In our weekly L10 leadership meetings, there's trust like a family. When I argue with my family, I never worry about what they'll tell others. We never worry that they would leave us because of some mistakes. They may scold us and call out our shortcomings. But we know they will never leave us. That trust is there. When we acknowledge our shortcomings, others accept their own parts too. It also empowers others to freely share their shortcomings. A strong, family-like bonding is created.
Leading by example is very important. If you're not doing something yourself and expecting someone else to, that's not the right expectation. Many times we come to the office at a certain hour but expect everyone else to arrive earlier. I don't think that is a good strategy. If you are coming to the office at 2 pm for some reason, you also need to accept the fact that others might do the same. Integrity is important—am I doing what I'm saying, and am I saying what I'm doing? Lack of integrity erodes overall trust.
Every learning is valuable. As technology people, we might dismiss non-technical learning, but every learning outcome can lead to better results. Sitting with the finance team, even if finance isn't my core, helps a lot in strategic decision-making because you grasp the intent behind everything.
Finally, my lesson is that we need to be open, flexible, vulnerable, and honest. Once we can do that, there's nothing to lose. Yes, brother, I make mistakes. I often make errors. I also do some right things. It's not about obedience but respect, which needs to be there. Otherwise, things tend to fail.
Ruhul: This is beautifully put. You can call them meta principles. A broad approach to look at the world and the work, and a way to operate. For today, one final question: your favorite books.
Mizan: I read mostly fiction until around 2006. I went to Dhaka College, and access to books was better around Nilkhet. I had a huge collection of fiction. As I entered the professional world, I started reading books related to my work. I came across books like Rich Dad, Poor Dad, and How to Win Friends and Influence People.
More recently, I've started looking after our sales team, and as part of that, I am studying sales a bit. I am currently reading this book called Sales Playbook for Hyper Sales Growth.
Raisul heavily influences my reading, particularly from a professional perspective. If he finds something interesting that seems helpful, he shares it with us all.
If I name only a few books, How to Win Friends and Influence People is an all-time favorite of mine. I think everyone should read it.
Similarly, I've read many sales books, but the one I am reading now, "Sales Playbook for Hyper Sales Growth," is one of the best. The processes are explained beautifully. Many sales books are abstract. This one clearly tells you what to do and why. If a customer says, "We'll let you know later," instead of just leaving, the book suggests asking about their needs on a scale of one to ten. If it's low, don't push; if it's high, ask what it would take to make it happen. This might seem simple, but it's very effective. Similarly, it offers excellent advice on approaching customers who say your service is expensive. Anyone who reads it will relate and be impressed.
Another book I would mention is The Elephant Catchers by Mindtree co-founder Subroto Bagchi, which is about how they grew from a small company. It talks about how they sometimes had to reject or fire clients who were bottlenecks for their potential growth, even if they were big revenue streams. There are many books; these are some of the names that come to mind.
The first lesson is we all make mistakes. Making mistakes is alright. The important thing is whether we learn from those mistakes. When we put our mistakes to better use by learning from them, that is the best way to learn anything meaningful. When we allow ourselves and others to make mistakes, we empower ourselves to get things done.
Ruhul: Excellent suggestions. I think this is a good place to end today's conversation. Thank you for being generous with your time and insights. This has been an enlightening and educational conversation for me.
Mizan: It has been an interesting conversation. Thank you for having me.