Intech Solutions are graph based fraud detection developers for companies and governments throughout Australia and New Zealand.
Graph fraud detection
Intech Solutions are the exclusive distributors of TigerGraph throughout Australia and New Zealand. TigerGraph is a graph database platform used by large enterprises and data-driven organizations that need to analyze complex, interconnected data.
Australia & New Zealand
Discover how Intech Solutions assisted the Australian Tax Office in implementing graph technology to aid in curbing tax avoidance and fraud.
Australian Financial Review
In the fiscal year ending June 2024, Australia experienced $868 million in card-related fraud.
Notably, 92% of this fraud—amounting to $785 million—was attributed to Card Not Present (CNP) transactions, which include online purchases and digital services.
This represents a 29% increase from the previous year.
Sophisticated synthetic identities are used by modern scammers to create accounts that mimic authentic behaviour by fusing real and fake data. These threats are difficult for traditional fraud detection systems, which are based on relational databases and concentrate on the behaviour of individual entities. These days, effective detection necessitates examining relationships over time across networks of accounts, frequently incorporating third-party data—something graph technology is specifically designed to manage.
Consider healthcare fraud involving treatment facilities for opioid addiction.
Traditional systems may overlook suspicious activity in healthcare fraud detection, such as a doctor routinely sending patients to one opioid treatment facility. TigerGraph’s deep link analytics are used to find hidden connections, like a shared past address between a treatment centre administrator and a doctor, using third-party data.
Through the integration of external sources and numerous data hops across patient claims, TigerGraph exposes collusion that relational databases are unable to readily identify because of slow joins and rigid schemas.
Fraud detection must be done quickly because delays cause businesses and their clients to suffer increasing losses. TigerGraph’s real-time detection feature allows for immediate action. For instance, China Mobile, which serves more than 900 million users, analyses call patterns from prepaid subscribers in real time using TigerGraph. In order to minimise impact and maximise response time, suspected fraud calls immediately result in alerts or redirection to their call centre for investigation.
Since less than 1% of financial transactions, medical claims, and phone calls are fraudulent, machine learning models have little verified fraud data to work with. This problem is solved by TigerGraph’s native parallel graph database, which analyses intricate relationships in real time. For instance, TigerGraph at China Mobile separates reputable clients from potential scammers by analyzing subscriber interactions over time to produce over 118 features per phone.
Graph database analytics for investigations, compliance and fraud detection
Terry Goodman, Managing Director of Intech Solutions, delivered a keynote address at the 7th Annual Australian Government Data Summit held on the 23-25th March 2021 in Canberra, Australia. Terry’s presentation covered an exciting new topic regarding graph database analytics and its use in investigations, compliance and financial fraud detection solutions. His presentation included:
- Relational vs key-value vs graph databases
- The next generation of graph databases
- Preparing data for graph database analytics
Terry also shared an interesting case study, describing the Australian Tax Office (ATO) adoption of the TigerGraph graph database. TigerGraph is a modern graph solution winning global applause for its unrivalled speed and scalability.
Why graph databases are your best solution to fraud detection
- Uncover Hidden Relationships: Graph databases naturally model complex connections between entities—like people, accounts, devices, or transactions—making it easier to detect suspicious patterns and collusion that traditional databases miss.
- Real-Time Anomaly Detection: Graph queries can traverse multiple hops quickly, enabling real-time detection of unusual behavior across connected data, such as sudden spikes in activity or indirect links to known fraudulent actors.
- Adaptive & Scalable Modeling: Fraud tactics evolve constantly. Graph data models are flexible and schema-light, allowing teams to rapidly adapt and extend their detection strategies without costly redesigns or downtime.
Talk to our graph fruad detection specialists about your project.
Speak to our graph fraud detection specialists /consultants about how your organisation can leverage Intech’s ultra-fast graph database solutions.
Our graph fraud database experts can help you to successfully implement your project
Leverage our graph fraud database experts to implement your project. Intech Solutions has over 20 years experience delivering high performance solutions, on time and on budget.
Intech’s IQ Office solution was embedded into the Databank software used by our clients to prepare and encrypt their data prior to upload. This led to a material reduction in risk for Westpac Databank by reducing the potential for formatting errors in the data preparation process and therefore maximising match rates.
Westpac Databank
Sydney
We are really glad we chose Intech as our partner. The project was complex with short timelines, high data volume and critical performance requirements, yet Intech handled each step well – delivering an excellent outcome for us and the Insurance industry as a whole.
illion
Australia
TigerGraph is a graph database platform used by large enterprises and data-driven organizations that need to analyze complex, interconnected data.
