Case Study: Actis
- Trained AI to consolidate complex, multi-lingual signals
- Ability to tune the algorithm
- Created a highly visual dashboard
Actis is a global investment fund specialising in renewable energy. They work in a number of developing markets and are looking to enhance their predictive capability around supplier performance and risk. Orthodox company data sources tend to be backward looking and less reliable in emerging markets. Forestreet were asked to demonstrate the potential for AI to cut through the ‘noise’
Because the project, by definition, was looking at inconsistent data sources we decided to avoid any ‘black-box’ components for the AI. Instead we decided to focus on changes in signals and allow the experts in the business to tune the algorithm based on their substantial experience. The AI is trained to consolidate complex, multi-lingual signals into ‘events’ and understand which are relevant but everything is surfaced through a highly visual dashboard so that the analysts retain control and confidence.