To Infocom and Christos Kotsakas
Technology has decisively moved from the background to the decision-making table. From Artificial Intelligence and data to cybersecurity, the cloud, and critical infrastructure, the choices made by technology leaders determine how an organization grows, manages risk, and responds to change. In this “Tech Leaders” feature, institutional representatives, experts, and market executives map out this new reality. Their insights highlight the IT Director as the vital link connecting innovation, management, and measurable business value!
Antonis Psaras, Managing Director of Microbase, shares his expert perspective on the importance of balancing the coexistence of technologies in AI projects.
The rapid evolution of AI poses a significant challenge for those working on related projects, as the technology’s maturation and adoption rates shift at breakneck speed—making project management exceptionally difficult.
We experienced this firsthand during the development of the Unified Citizen Service Center—part of the Greek Public Sector’s “Unified Digital CRM Infrastructure” project, in which Microbase participated. We were called upon—and succeeded—in leveraging new AI capabilities without compromising the project’s stability, security, or reliability, all while continuously revising the design.
Such scenarios require the use of NLU, Generative AI, and LLM technologies. However, the bar is set high, as the optimal architecture relies on the right combination of these technologies—a truly complex equation.
NLU (Natural Language Understanding) platforms represent a “mature” technology. They enable the comprehension of user intent and offer a structured approach to service development, integration with telecommunications infrastructure, and the ability to create controlled business workflows. They are particularly effective in real-time applications.
Conversely, solutions based on Generative AI and LLMs (Large Language Models) possess impressive capabilities for understanding and generating natural language, offering rapid development, more natural communication, and the ability to cover a vast knowledge base. However, their predictability is lower, which can create challenges in scenarios where accuracy, compliance, and consistency are critical requirements.
In the realm of contact centers and voice/chatbots, the future does not belong exclusively to LLMs but to hybrid models where NLU leads and guides the AI. This approach delivers the best of both worlds: structure, control, and reliability on the one hand, and natural interaction and intelligence on the other. When combined with support from human agents, this ensures high-quality service delivery as well as customer and employee satisfaction.