After years in the structured world of AWS, Tomasz Stachlewski has taken the helm of technology at Comarch. In his latest interview, he explains why he traded a global giant for a Polish organization undergoing its biggest transformation in years. How does he plan to use his cloud expertise to make ERP intuitive, why is he introducing a PR/FAQ culture, and how will Comarch technologically challenge Silicon Valley?
You changed colors after many years at a global giant (AWS). Why Comarch and why now?
Comarch is at a very exciting turning point—it’s undergoing a massive technological and organizational transformation. It is rare, both in the Polish and European markets, for such large-scale changes to be limited to mere announcements; here, they are happening in real time, week by week. That was the deciding factor for me. After years at AWS, where we operated at a massive scale but in a highly structured, predictable environment, I was looking for a place where I could have a direct impact on product direction and the entire organization. This company has solid technological foundations but immense potential to accelerate. My goal is to co-create an organization that will be one of the most dynamic tech firms in our region—not just in scale, but in how we build products.
As CTO, what is your absolute priority for the first 180 days?
The main priority is understanding the organization and its people. Comarch is a complex organism; good technological decisions can only be made when you truly understand how it functions in practice. That’s why my first months are about listening and building a full picture. The second key area is Artificial Intelligence—taken from a very pragmatic perspective. On one hand, we want to accelerate its use in products to make them more modern. On the other, we want AI as a tool for our internal teams to work faster and more effectively.
Moving from AWS to Comarch is a shift from a Cloud Service Provider (CSP) to an Independent Software Vendor (ISV). How will your experience influence Comarch’s tech stack optimization?
This is one of the most important things I bring to the table. At AWS, I worked with hundreds of companies, from startups to enterprises. I saw diverse approaches to architecture and costs. One thing is constant: any functionality can be built in multiple ways. The difference lies in which approach is the fastest, safest, and most cost-effective. In today's world, those who deliver value quickly while maintaining stability win. My goal is "conscious architectural design"—systems that are scalable and allow teams to move fast without bloated costs.
Many still associate ERP with "heavy" on-premise systems. What is your recipe for making ERP "sexy" and modern?
ERP must stop being a data-entry system and become a decision-support system. This means great UX, real-time data access, and AI as a natural support layer. It’s also crucial that everything is accessible via browser—that is the natural interface today’s users expect. A modern ERP shouldn't require the user to know the system; the system should understand the user.
Cloud-Native is more than just containerization. Do you plan a deeper adoption of Managed Services and Serverless?
In my vision, Comarch is opening up significantly more to external solutions, including cloud-based ones. If managed services or Serverless allow us to build faster, deliver value sooner, and increase security, we should use them. It’s not about blindly following one model, but about making conscious technological choices rather than sticking to the status quo just because it’s familiar. In today's world, the lack of change is the greatest risk.
Is SaaS the only way forward for business systems, or do you see room for hybrids?
SaaS is dominant and often the best choice for fast growth. However, I don't believe in a "one size fits all" approach. The business world is complex—there are regulated industries and specific security requirements. Ultimately, the hybrid model wins because it is flexible and tailored to the client's needs.
You often speak about the "democratization of technology." Does this include Low-code/No-code platforms within ERP?
Democratization means making modern solutions accessible to everyone, regardless of their scale. We are open to Low-code or No-code where it brings value, allowing clients to build their own extensions without interfering with the core system. However, we must maintain consistency and security so that client-side development doesn't compromise system stability.
Where do you see the most practical use for Generative AI in ERP?
I see the greatest value in automating manual, repetitive processes—such as handling invoices, contracts, and accounting entries. Secondly, interaction: instead of searching for reports, a user should simply ask the system in natural language. Thirdly: decision support—identifying anomalies and providing recommendations based on existing data.
How do you avoid the "AI for the sake of AI" trap? How will you measure innovation?
We are introducing the PR/FAQ approach. Before we build anything, we describe to the client what the feature looks like and what problem it solves. If we can't clearly articulate the value in a document, we shouldn't build it yet. We are also developing an internal telemetry platform, GALILEO, to understand how our systems are actually used. We want to move from intuition-based decisions to data-driven ones.
How do you intend to compete technologically with Silicon Valley giants?
My ambition is for Comarch to be an organization that can be compared to the biggest players in terms of tech quality and development pace. But I don't believe in carbon-copying Silicon Valley. Our strength is the combination of solid product foundations, deep domain knowledge, and proximity to European clients. We must play at the highest level—simplifying architecture and utilizing AI practically—but stay better suited to the specific markets we serve.
How will you automate compliance and localization for global markets?
The key is architecture. We are building modular, configurable systems where local differences are managed as sets of rules and parameters rather than being "hard-coded." AI will also play a larger role in interpreting regulatory requirements, allowing us to scale globally faster and cheaper.
What is your framework for dealing with technical debt?
My principle is: long-term success over short-term success. I’m not a fan of rewriting everything from scratch just because the old system isn't "perfect." If an area can be refactored, we do it. If it starts slowing down the business or generating too much risk, we use the Strangler Fig Pattern (gradual replacement). The goal is an environment that lets teams act effectively, not just today, but years from now.
Will you continue your role as a technology "evangelist"?
Definitely. Sharing knowledge is a passion of mine. A CTO shouldn't only look inward; it's vital to participate in the market discussion and educate the ecosystem.
What do you do to stay "technologically sharp" amidst management duties?
I believe managers must stay technical. I spend a lot of time on AI—not just reading, but hands-on experimentation. In AWS, I kept up through certifications; today, I dive deep into models and tools. This allows me to actually advise my teams rather than just managing them from a distance.