AI output evaluation: cross-company workshop

Dear Software Center members,

You are invited to a cross-company workshop on AI output evaluation.

We are all busy thinking about or already applying AI assistants and agentic AI workflows to improve and automate every applicable process in our companies. But how do we know if it’s good enough? Which AI model works best? How to compare them, how to choose? How to monitor and control the quality over time? This is a fundamental problem for any application of AI in processes, and what often keeps us from maturing from prototypes to production.

In this workshop we invite you to:

• Share your experiences, challenges and solutions
• Learn from Chi Xiao about current state of the art in practically evaluating AI performance
• Together identify and discuss top opportunities, challenges and topics of interest

Chi Xiao is a developer at Ericsson who specializes in requirements engineering and the application of AI in that field. She recently published at ICSE 2025 on UML Sequence Diagram Generation using AI.

You are welcome attend in person at Ericsson in Linköping, Datalinjen 3, or via Teams.

Please forward to anyone in your organization to whom this may be of interest, and RSVP to this invite by Jan 7 so that we may gauge interest and plan accordingly.

Best regards,

Daniel Ståhl, Chi Xiao and Jan Bosch