Interview: How to get started with Agentic AI?

Maria Kern

Most companies don’t lack ideas; they lack clear direction. They ask themselves: Where do we start? Which use case is truly worthwhile? And how do we avoid getting stuck in a pilot phase? At first glance, Agentic AI seems complex because it affects processes, data, systems, and organization all at once. This makes a structured approach all the more important—and it doesn’t start with the technology, but with specific business goals.

Maria Kern

The AI Workshop is deliberately designed so that companies can not only understand agentic AI but also experience it firsthand. In a compact format, business units and IT work together on real-world challenges—for example, in marketing, customer service, sales, e-commerce, or data management. The goal is not to develop a theoretical roadmap, but to achieve a tangible result: a first functional AI agent tailored to a real-world use case, including measurable KPIs, concrete implementation steps, and a concept for orchestration and seamless integration into existing systems. Participants experience the entire process, from the initial idea through the agent design to an initial production-ready workflow.

Maria Kern

We always start by asking: Which business goals currently have the highest priority? These might include improving efficiency, shortening time-to-market, personalization, or improving conversion rates. Building on that, we work together to assess data and system readiness: What data is available? What systems are in use? Where are there gaps? We also take governance requirements into account early on to ensure that data flows, responsibilities, and control mechanisms are clearly defined from the start. Only then do we look at the potential and possible use cases for agents: Which tasks are repetitive or time-sensitive? Where can an agent prepare or automate decisions? Which types of agents make sense – such as analytics, content, personalization, or orchestration agents? In the subsequent build phase, the teams – supported by valantic – develop an initial production-ready agent tailored to the prioritized use case. We ensure that the agent is not only technically functional but is also embedded into existing processes – including control mechanisms and guidelines that ensure trust and transparency.

Maria Kern

From my perspective, it’s above all about commitment. Many approaches end with ideas, sketches, or presentations. We, on the other hand, build a tangible solution – one that takes real-world processes and goals into account – often orchestrated across multiple systems and teams. An important part of the workshop is integrating the developed agent into everyday work. We define which process it supports, how its impact will be measured, and who is responsible for it. This leads to the next implementation step: a clearly defined development sprint that enables the agent to transition from prototype to production.

Maria Kern

The workshop is particularly well-suited for companies that are moving beyond the phase of isolated AI experiments and want to systematically integrate Agentic AI into their organization. Companies with interdisciplinary teams or more complex structures benefit the most: departments such as marketing, sales, IT, and operations can work together toward a clear goal. A major advantage of this format is that we tailor it individually to the specific requirements and circumstances of each team. The format is flexible enough to address the respective structures, challenges, and goals.

Maria Kern

Companies walk away from the workshop with three key takeaways: First, clarity on which Agentic AI use cases are truly relevant to them and create the greatest value. Second, a working prototype that is aligned with real-world goals and was co-developed by their own teams. And third, a concrete implementation roadmap that shows how the prototype will make an impact in production. This transforms Agentic AI from an abstract vision of the future into a manageable component of the company’s own organization. Companies with clear governance and implementation plans are significantly more successful at scaling AI solutions and generating business value. This is precisely what the valantic AI Discovery Workshop aims to achieve. This results not only in the creation of an initial agent but also in a shared understanding of how Agentic AI can be sustainably established and scaled within the company.