In December 2025, KMS Technology — a U.S.-based Digital Engineering, Data, and AI services company headquartered in Atlanta — acquired Addepto, a leading AI and data consulting firm headquartered in Poland. In this exclusive interview, Leo Tucker – CEO of KMS Technology and Artur Haponik – CEO of Addepto, discussed the rationale behind the partnership and how the combined entity is delivering greater end-to-end value to global enterprises adopting AI at scale.
Leo Tucker: Addepto needed to complement its strength in data and AI with broader digital engineering services, while KMS Technology needed to deepen its data and AI capabilities. Together, we now provide a more complete, end-to-end suite of services with significantly stronger data and AI expertise.
Artur Haponik: Exactly. By combining our strengths, we can deliver AI and Big Data projects from ideation through implementation, seamlessly integrating both AI components and the supporting software – something that would be far more difficult for either company to achieve independently.
Leo Tucker: The biggest learning has been just how complementary our capabilities truly are. While we recognized the capabilities fit during due diligence, the level of synergy has proven to be even stronger than we originally anticipated.
Artur Haponik: Honestly, the biggest surprise was the cultural alignment. It’s remarkable — our core values and missions are exactly the same, just described in different words. We’re definitely on the same page.
Leo Tucker & Artur Haponik: The biggest challenge remains creating AI solutions that are genuinely useful and deliver tangible benefits to end users. As a combined organization, we place strong emphasis not only on delivering solutions to our clients, but also on ensuring these solutions create real value for end users. Once a solution is useful, the next major challenge is ensuring that AI results are reliable, consistent, and free from excessive errors or hallucinations.
Leo Tucker & Artur Haponik: A typical client engagement begins with understanding the client’s business goals and identifying opportunities where AI can deliver real impact. We start by defining clear objectives, gathering and analyzing data, and designing a solution that aligns with those objectives. Next, we move into development and iterative testing, ensuring the AI solution is robust, reliable, and adds tangible value for end users. Finally, we support deployment and continuous improvement, monitoring the solution’s performance and making adjustments as needed.
What differentiates working with the combined KMS Technology and Addepto team is our fully integrated approach. Instead of coordinating between multiple teams, which can create delays, miscommunication, or misaligned priorities – our partnership combines domain expertise, AI development, and engineering execution under one cohesive workflow. This ensures faster delivery, higher quality, and solutions that are both technically sound and directly aligned with business outcomes.
Leo Tucker: I believe that many software engineering firms have been scrambling to add AI capabilities in the last year or so. KMS Technology chose to acquire Addepto in part because of their deep and established experience – Addepto has been working with AI for 8 years. Their level of depth and experience differentiates KMS Technology from other software engineering firms that are jumping into AI in the last couple of years.
Artur Haponik: What sets us apart is our deep expertise and proven track record in AI delivery. At Addepto, 97% of our team are AI engineers, and our founders bring extensive experience delivering enterprise AI solutions. Combined with KMS Technology’s software engineering excellence, this partnership enables us to deliver end-to-end, high-quality AI and digital solutions that are both technically robust and business-focused.
Leo Tucker & Artur Haponik: The main reason AI initiatives fail is a lack of real business value. Many proofs of concept don’t achieve the expected benefits and are eventually abandoned. Success requires more than just writing code or implementing an AI model – it’s about deeply understanding the business problem and designing solutions that actually solve it.
Specifications rarely capture all the challenges we encounter in data and algorithms, which is why it’s crucial to stay business-value-oriented and be ready to pivot the solution as needed. Equally important is testing early versions with end users and incorporating feedback, ensuring the final product delivers tangible, measurable value to customers.
LOOKING AHEAD
Leo Tucker: The level of integration will continue to tighten. While we are already working well together and seeing synergies across our offerings, over the next 12-24 months we will naturally integrate more tightly and uncover additional ways that we can accelerate together.
Artur Haponik: We believe this partnership will create exponential value – where 1 + 1 = 4. We will continue investing in deepening our AI and Data expertise, particularly through our Databricks partnership. We also plan to work even more closely with clients globally to help them accelerate their internal AI initiatives.
Leo Tucker: The most important starting point is the business problem itself. Pursuing AI for its own sake rarely leads to success. Organizations must focus on clearly defined business challenges. With that clarity, our teams can develop solutions that deliver meaningful and measurable outcomes based on our experience and capabilities.
Artur Haponik: It’s also critical to partner with organizations that have deep AI expertise and relevant industry experience. Experienced partners can provide valuable insights into what works, potential risks, and realistic expectations. This guidance significantly increases the likelihood of success and provides practical market knowledge that’s difficult to gain through trial and error alone. By starting with informed guidance, organizations can scale AI responsibly and create real business impact.
Category:
Discover how AI turns CAD files, ERP data, and planning exports into structured knowledge graphs-ready for queries in engineering and digital twin operations.