Accelerating Supply Chain Decisions with Natural Language Queries
Supply chain buyers and manufacturing controllers often struggle to extract immediate answers from complex ERP systems, leading to workflow bottlenecks when identifying late purchase orders that threaten upcoming production runs. According to a November 13, 2026, report by ERP News, Databricks has introduced a Genie plugin for ChatGPT designed to address this operational challenge by allowing employees to query SAP ERP data using natural language.
For organizations evaluating conversational AI for their manufacturing data, the underlying architecture requires several prerequisite integrations. Based on the Databricks November 13 update, the route into ChatGPT builds upon an integration between SAP Business Data Cloud and Databricks. Administrators use SAP Business Data Cloud Connect to share selected SAP data products into an existing Databricks workspace. This is achieved through zero-copy sharing via OpenSharing, making the data accessible and governed through the Databricks Unity Catalog.
Translating Manufacturing Data into AI Answers
Connecting the systems is only part of the implementation process; the AI must also accurately interpret specific industry and company terminology. The Databricks Genie plugin relies on the Genie Ontology, which became generally available on November 9, 2026. This ontology supplies the business definitions necessary to understand complex queries. When a manufacturing controller asks to trace a margin variance, Genie uses these business definitions to run the analysis natively in Databricks, ultimately returning the results alongside direct links to the underlying data sources.
Manufacturing decision-makers looking to implement similar capabilities should ask the following evaluation questions:
- Are our critical supply chain and ERP data products accessible via a governed data layer like Unity Catalog?
- Have we clearly defined our internal business metrics (e.g., margin variances, production schedules) in a structured ontology?
- Does our current data architecture support zero-copy sharing to minimize duplicative data pipelines?
Preparing a manufacturing organization for AI-driven operational insights requires strong foundational data architecture and process mapping. Whether your organization is looking to streamline complex enterprise data sets or needs comprehensive AI and data intelligence services to prepare for conversational analytics, establishing proper governance is the critical next step. To discuss how to prepare your supply chain workflows for advanced data querying, contact Decision Intelligent to review your implementation prerequisites.