IsraelManufacturing2026 – ongoingHP Indigo
Prediction model for manufacturing yield
HP Indigo raised yield on one manufactured part from 60% to 92% with a prediction model
Problem
HP Indigo relied on separate legacy data warehouses, manual spreadsheets and fragmented ERP and manufacturing systems. Decisions were slow, and production sometimes stopped while teams traced parts. The company managed more than 3,500 data volumes, about 10 terabytes.
Approach
HP Indigo moved its legacy data to Databricks and built a prediction model for one manufactured part to tune production line parameters. It used Unity Catalog lineage to trace consumables from manufacturing to customer sites. It also added Databricks Genie so business users can query governed data in plain language.
Results
Takeaway
The yield gain covered one manufactured part. The company did not say whether it plans models for other parts.
Categories
- Size
- Large enterprise
- Industry
- Manufacturing
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
Databricks reports that HP Indigo built a prediction model for one manufactured part and raised yield from 60% to 92%. Consumable traceability fell from 3 days to 1 hour with Unity Catalog lineage. The story also mentions natural-language data access through Genie, with no figure.
Full details are in the original sources.
