Decision Modelling

Challenge: In a supply chain of goods, the agent handling the containers wanted to predict which containers were likely to have poor quality products to avoid checking all containers that arrive in the harbour. The Agent thought the data was not available, so set out to collect data first for a period of 2 years before running an analyses project.

Solution: Just a quick scan with Decision Modelling already revealed that the data the agent was aiming to collect was neither sufficient nor focused towards the decisions he wanted to make (it did not include collecting the relevant social media content, for example). Decision modelling delivered proof of a different approach and much earlier start of the predictions. SuperGraph Predictive Analytics could be used as early as 2 months into the data collection and learn to adapt to new trends and patterns on the fly.

Benefit: In this case the benefit was 2 fold. Significant costs were saved on a 2-year data collection project, which never would have delivered sufficient data to design reliable models. And secondly, the first predictions were made as little as 2 months into the project, delivering immediate cost reductions and a Return On Investment 18 months earlier.