Leveraging Natural Language Processing for Tax Update Notifications
Implementation Time:
9 months
Solution Provider: AI Singapore
GIC is a sovereign wealth fund established by the Government of Singapore in 1981 to manage Singapore’s foreign reserves. Its mission is to preserve and enhance the international purchasing power of the reserves, with the aim to achieve good long-term returns above global inflation over the investment time horizon of 20 years. It has a network of 10 offices in key financial capitals around the world.
- Every country has their own tax code and the codes change frequently
- Changes in a country’s tax code have implications across the multiple investment portfoliosÂ
- Tax team keeps abreast of these changes and provides consolidated advice to their stakeholders
- Tax team’s analysts receive different reports from multiple sources – confusing and tedious to track
- They have to manually screen these updates for the topic and region of interest
How can AI help the tax team have a more streamlined set of data and reports?
Developed an AI system using an ensemble of models that automatically classify the reports based on topics and country of relevance:
- AI solution streamlines initial review so the team can focus on value-added tasks e.g. layering information with expert insights
- The system also includes a feedback loop that improves the model over time
Outcomes
- AI model achieve analyst-level performance in classification
- Passed User Acceptance Tests with 82% relevancy
- Models has been adopted by GIC and deployed for internal use
PDF Document
This is the last content of this tab. If you do not see any resources above, it means the solution provider have not provided any resources. Feel free to contact Solution Provider for more information.
Implementation Time
9 months
Use Case Brochure
