A large Finnish international industrial company is building a semantic layer that brings together SAP, Snowflake and document data for use by language models and agents. For example, an agent could explain an anomaly spotted in reporting by retrieving context from background sources. The work starts with an MVP, most likely in the sourcing area.