As the manufacturing industry continues to embrace digital transformation, the integration of Language Models (LMs) presents both significant opportunities and notable challenges. This talk explores how LMs can be applied to real-world manufacturing—from knowledge retrieval to automating documentation and reporting. But the journey is not without obstacles. We'll examine the unique challenges of deploying LMs in industrial settings, including domain-specific language, hallucination risks, latency constraints, and integration with legacy systems. Drawing from practical use cases and lessons learned, we’ll explore strategies to overcome these hurdles—such as fine-tuning with manufacturing data, retrieval augmented generation, and edge deployment. Attendees will gain a grounded perspective on how to effectively use LMs to drive efficiency and innovation on the factory floor.