Problem solving is a universal process faced by all organizations when dealing with issues, complaints, incidents, deviations or non-conformances.
BPA’s development lab is currently exploring AI for predictive analysis applied to problem solving processes.
The developed machine learning algorithm compares similarities between new entered incidents and historical data, and suggests root causes, corrective actions for a quick and efficient solving.
The used algorithm was trained on the entire Wikipedia database for optimal similarity searching, not only looking for exact matches but also analogous matches that have the same meaning by using comparable terms, expressions and synonyms. For instance, when reporting a new incident “customer service was not good”, the algorithm will propose 3 root causes: lack of communication skills, lack of domain knowledge, lack of responsible behavior. In this case, the algorithm found a similarity with a previous incident “customer service delivered poor response”.
Behind the scenes, the process will compare new incidents with previous ones and calculate a correlation factor for each item. Suggested actions and root causes corresponding to the highest correlation factor will be proposed to the investigator.
The data model is updated each time new data is entered in the BPA software, improving the algorithm’s efficiency.
BPA’s AI features will radically help organizations solving problems faster and more efficiently, reducing costs and improving customers’ satisfaction.
BPA will progressively introduce AI to suggest risks and actions related to any quality improvement process, like audits, inspections, objectives, indicators, swot analysis, etc.
Conversational intelligence – read article and view video here – and AI features will soon be available with BPA Quality & Risk Management on Office 365.
BPA is proud to be on the front line implementing AI for regulatory compliance.