The client was not able to accurately predict case resourcing needs and case profitability using incumbent, inefficient manual processes.
A machine learning platform was built that:
- Implemented predictive models built on historical time entries by position in the firm, used text analysis to complete missing and incomplete data, and employed clustering algorithms to group cases together with similar resource profiles
- Analysed historical invoices to build the case profiling platform
- Created a quick access dashboard that was made accessible and readable by all levels of technical skill
The machine learning platform provided accurate forecasts of case resourcing needs and profitability, as well as the ability to analyse trends in hours worked by role and phase of case in addition to trends of case duration.