The ATO is building a big data platform to tackle tax evasion

The Australian Taxation Office is building a network analytics solution called ‘ANGIE’ to help its tax avoidance taskforce discern complex, multi-layered relationships between clients.

The solution, which will be underpinned by a graph database, is being developed as part of the taskforce’s data and analytics program.

The taskforce was setup in 2016 to crack down on tax avoidance by multinationals, large public and private businesses and associated individuals.

ANGIE - or the Automated Network & Grouping Identification Engine - will automatically identify and group clients to understand their relationship to one another.

It will allow the taskforce to detect “patterns of interest” and visualise new links between clients.

A “critical component” of ANGIE, as well as the ATO's wider network analytics and intelligence capability, will be a graph database, which the agency this week opened tenders for.

A graph database is used to represent complex networks by storing data about entities (nodes) and their relationships (edges).

The approach to market follows advice from strategy consulting firm McKinsey, which completed a “high-level end-to-end design of ANGIE” last year.

McKinsey found that “a native graph database would enable significant processing and performance speed, capturing more detailed information … across the organisation”.

The graph database, which is a first for the ATO, will allow the agency to capture, visualise and analyse “large and complex networks of relationships”.

Structure data will be pulled from the ATO’s Teradata enterprise data warehouse, while semi-structure and unstructured data will be sourced from its Cloudera enterprise data hub. 

The the graph database is also expected to be used by the global tax enforcement group J5 to “discover enablers of tax time”. The group is made up of revenue authorities from the US, UK, Australia, the Netherlands and Canada.

The ATO expects the cloud-based COTS graph database to be delivered from July in order to meet the project timelines for ANGIE.

“The ANGIE project timelines call for prompt integration of a graph database into the ATO - Q3 of this calendar year,” tender documents state.

It may also need to procure IT services to “assist with the installation and integration” of the graph database, though this will depend of the product selected.

The agency expects to award a five-year contract in May.