This talk presents Bosch’s structured approach to developing an AI-based visual inspection system. Starting from an internal modeling competition, the process emphasizes clear objectives, consistent datasets, and model logging for transparent evaluation. Collaboration with clients helps define practical performance metrics. The transition to production focuses on robust orchestration, cost efficiency, and inference speed, ensuring that the chosen model is both technically sound and deployment-ready.

Bosch shares its experience in unifying supply chain data using an end-to-end platform built on Databricks. The session outlines how the integration of dbt and LLMs enhances data transformation and insight generation. Key takeaways include platform interoperability, architectural best practices, and how these technologies drive efficiency in complex supply chain operations.