VTNZ came to Provoke with an exciting data science challenge: could we use their comprehensive vehicle inspections data warehouse and data pipeline to surface vehicle condition and maintenance insights for their customers?
Provoke worked with VTNZ and Microsoft to identify some high-value business and customer problems to address, and then worked through the data available to see what may be possible. After performing regression and time series analysis, we concluded the data was well suited to tackle predictive maintenance and operational efficiencies through machine learning. Provoke took a Proof of Concept approach to quickly demonstrate possible results and potential business value.
Through the application of AI technology, Provoke developed a state-of-the-art predictive maintenance solution for VTNZ that conceptually proved VTNZ customers could check a car model for common mechanical problems before purchasing or submitting it for a warrant inspection.
We also developed a PoC model for operational efficiency of VTNZ stations, enabling management to effectively plan for busy periods.
These results provide VTNZ with the opportunity to further develop a platform for driving better customer engagement and improve business performance.
Provoke delivered the solution as a self-contained web application and API module using Docker technology, making it highly scalable and easily deployable in any environment