Dr. Marcia Gottgtroy 
Inland Revenue New Zealand - Principal Intelligence Analyst – Knowledge Engineering 

A Bachelor in Computer Science when Artificial Intelligence was in its gold days and an exposure to challenging decision management projects led Marcia to develop a passion for BRMS and Decision Systems, to a M.Sc. in Artificial Intelligence and a D.Sc. in Engineering. During the last 25 years or so, she has been involved and led the development of Decision Management Systems solutions in deep sea technology, logistics, agriculture and tax lately. Marcia has always been a true believer and practitioner of knowledge in action through the application of knowledge engineering principles and hybrid solution approaches.
The Compliance Management Environment (CME) is a decision management platform in development at Inland Revenue New Zealand (IRD) to support the strategy of transforming IRD in a “21st Century Tax Administration Department”.

The core component of CME is a decision engine that makes use of competing models to suggest “the best next action” based on events related to customers and their interactions with IRD.

The platform uses a combination of statistical, machine learning, semantic and rules-based applications, and human intelligence to reason over a holistic view of the events, entities, their associates and the communities in which they operate.

Knowledge engineering methodologies are employed to create a learning environment that captures all stages of the decision making process along with all data collected, inferred knowledge, risk profiles, models’ performance metadata and the decisions made by investigators interacting with the system. This collective body of knowledge is used to manage the decision making process.

This presentation will focus on Goods and Services Tax (GST), which has been in production since 2013. It will show that traditional risk management strategies focused on the GST return in isolation have proved inefficient in dealing with non-compliance. It will demonstrate that the diversity of customer profiles, seasonal/industry variance in businesses performance and the way companies are structured are key factors for compliance behavior risk assessment and in consequence decision management.