|Personal vehicle insurance||A top 5 insurance company wishes to improve the profitability of its vehicle insurance product while offering dynamic pricing to its current & new policy holders||Used past and present customer features and claims data to define clusters using a Machine Learning approach.||Under implementation|
|Financial Services||A large regional bank wanted to increase the purchase probability of three advertised financial products: loans, mortgages and credit cards.
||The team used a combination of Naive Bayes and Logistic Regression techniques to calculate propensity scores and estimate the prediction accuracy.
||Under evaluation. Expected success rate increase of 25% compared to previous method used.|
|Electrical Power distribution||The national electricity company aimed at reducing energy losses along the power generation, distribution and transmission network.||Data collected from power meters was used to instruct Deep Neural Networks.||Under evaluation|
|Product Distribution||A regional distribution company wanted a better planning method for routing of trucks for delivery of non-perishable goods.||Predictive analytics and clustering analytics was used to determine best time/location sequence.||The total average percentage reduction in travel distance was 59% and the total average percentage frequency reduction in visits was 15%.|
|Electrical Power distribution||The national electricity company wanted to accurately predict short term load.||Deep Neural Networks was used to determine load forecasting.||Under evaluation|
|Import/Export||A regional car sales company wanted to optimize vehicle import to ensure high demand SKUs are available and low moving SKUs are kept at a minimum||Exponential Smoothing was used to assist Inventory Forecasting.||Decrease in the average time spent on lot by 55 days whilst maintaining the probability of a vehicle existing on the lot at one.|
|Personal Insurance||A large regional insurance company wanted to increase the product attach rate for their existing clients.||Created a Recommender System using past product sales and client features to automatically determine most suitable product.||Under evaluation|