From buzzword to real support: what are AI agents really?

Markus Danhel clears up misunderstandings right at the start: AI agents are not remote-controlled science fiction robots, but highly developed, proactive software systems that can complete tasks independently, understand context, plan and find solutions themselves. In contrast to classic chatbots or large language models such as ChatGPT, which wait for human input, AI agents act independently and proactively. "The big difference to today's conversational AI systems is that AI agents can act proactively and support me in my processes - like a personal assistant that thinks for me."
Danhel does not see the future as a world in which AI agents replace people, but rather as intelligent assistants that relieve employees and allow them to focus on value-adding tasks.

Concrete use cases: From HR to customer service

What does the use of AI agents look like in business practice? Danhel provides numerous examples from different sectors:

- IBM Ask HR:
At IBM, over 280,000 employees use the Ask HR platform to handle typical HR tasks such as time recording, vacation requests or management tasks simply by voice. "I no longer need to call helplines or click through various backend systems. My AI agent understands my request and implements it - 94% of all requests are now processed automatically. That saves a huge amount of time and stress."

- Zurich Transport Authority (VBZ):
An e-mail butler was introduced here, which pre-formulates customer inquiries and suggests them to employees. "The AI agent helps to manage the flood of emails more quickly, especially at peak times. However, the human remains in control at all times - the agent provides suggestions that the employee checks and sends. This has improved response times by up to 20 percent."

- Legal Tech & Patent Examination:
The startup ABP uses AI to automate patent examinations.
"Imagine 170 million patents worldwide - that's 25 times the height of Mount Everest if you stack the documents on top of each other. With ABP's Patent Butler, AI can take over the initial analysis and summarization, massively speeding up the work of patent attorneys."

- Healthcare & nursing:
In Swiss retirement and care homes, AI agents are automating administrative tasks so that care staff have more time to look after people. "The more I enable employees to focus on their core competencies, the more the entire organization - and ultimately society - benefits."

Change management: people at the heart of AI implementation

The introduction of AI agents is much more than a technical project - it is a cultural change. "Change management is crucial. Employees have to be brought on board and creating acceptance takes time. A good example is a Swiss retirement and nursing home that has gained more time for care thanks to AI because administrative tasks have been automated." Danhel emphasizes that employee acceptance and understanding of the technology is a key success factor. Only if people recognize the advantages will they perceive AI agents as support and not as a threat.

Empowerment & skills: making employees fit for the future

Another key to success is the targeted empowerment of employees:

- Promote understanding of technology:
"A basic understanding of the technology allays fears and helps to recognize potential."

- Openness to new processes:
"Companies need to create space to develop new ideas and question existing processes."

- Developing talents:
"For younger generations in particular, being able to work with modern technologies is a motivating factor. If you want to retain talent, you have to give them the opportunity to get to grips with AI."
Danhel recommends forming interdisciplinary teams in which technical experts and technologists work together to develop new solutions. Agile methods and design thinking help to implement innovations quickly and practically.

Collaboration as a driver of innovation

For Danhel, it is clear that the best AI solutions are created when companies, technology companies and experts work closely together. "As IBM, we are a tech company, but we don't know every detail of every industry. That's why it's crucial to bring together experts from different domains and think openly together: where can we redesign processes, where can we create real added value?" He advises companies to simply get started, try out initial use cases and learn from their experiences - this increases confidence in the technology and makes real innovation possible.

AI agents as practical helpers and drivers of innovation

AI agents are no longer science fiction, but practical helpers in everyday corporate life. They relieve people of routine tasks, increase efficiency and allow them to concentrate on value-adding tasks. Collaboration, the right mix of technology and expertise - and the courage to try new things - are crucial to success.

"The key is to bring the right people together, think openly and simply get started. This is how real innovations are created - for companies and society."

---

About the podcast AI Talk
AI Talk is the podcast about artificial intelligence in Switzerland - with real movers and shakers, experts and decision-makers from tech, research, politics and business. Hosted by Stephan Lendi and Dalith Steiger-Gablinger, each episode of the podcast highlights new perspectives on innovation, responsibility and the concrete application of AI technologies. Whether ethics, regulation, talent development or deep tech - AI Talk combines personal insights with strategic foresight.

Listen, network & think along now: