MicroStep-MIS: Responsible AI for meteorology and safety-critical systems

MicroStep-MIS develops integrated monitoring, forecasting, and decision-support systems for weather-sensitive and safety-critical operations. Combining in-house engineering, scientific know-how, and long-term field experience, they deliver reliable solutions tailored to real operational needs. With an existing AI product already in use, MicroStep-MIS wanted to ensure they were approaching its deployment responsibly, in line with the EU AI Act and with awareness of the obligations it places on companies delivering products in the EU market.

Given MicroStep-MIS’s existing technical maturity, our collaboration was more technology-focused than a typical AI literacy engagement. We combined participation in our group trainings with individual mentoring tailored to the company’s specific needs and AI systems. As part of the group trainings, participants covered topics including:

Through individual mentoring sessions for MicroStep-MIS’s team, we provided a general overview of EU AI regulation and assessed the applicability of the AI Act’s obligations to the company and its systems. We also consulted on their internal directive for the use of AI tools, reviewed their documentation-based chatbot, and covered topics including the AI Act as a practical guide for companies and public administration, prompting for administrative efficiency, and how to prevent risks associated with employees’ use of generative AI. We also conducted an assessment of the company’s readiness to adopt AI technologies.

The collaboration helped MicroStep-MIS strengthen AI adoption across the company and bring a clear regulatory and responsible-AI perspective to its growing portfolio of in-house AI solutions. These range from radar-based precipitation nowcasting and satellite-based thunderstorm nowcasting to intelligent LLM-powered support for their product suite, among others.