The Client
A large, East Coast-based appliance retailer with a sophisticated service and repair operation focused on maximizing first-time-complete repairs and customer satisfaction. The company dispatches technicians to customer homes daily, and every missed part on a service call means a return visit, added cost, and a frustrated customer.
The Project
VisionWrights built a predictive parts recommendation model to support service preparation. The solution translates unstructured customer issue descriptions into likely parts needed for specific appliance models. It serves as the first step in automating a manual, multi-step process — reducing the time required to identify and stage parts before technicians are dispatched, without changing how service teams work in the field.
The model was built as an MVP to prove the concept and establish a clear path forward. Rather than attempting a full-scale automation project, we focused on demonstrating value quickly — showing that the approach could work with real data and real workflows before scaling.
The Result
The MVP demonstrated strong parts recommendations and established a clear path toward improving first-time-complete rates. The work surfaced specific data and process improvements needed to increase model accuracy, giving the client a practical starting point for automating service operations and scaling data-driven decision-making.
- ✓Strong predictive accuracy for parts recommendations from day one
- ✓Clear path to improved first-time-complete repair rates
- ✓Specific data quality improvements identified for increased accuracy
- ✓Practical foundation for scaling automation across the service operation