Prospa Propels Australia’s Small Businesses with Neo4j Graph Technology

Australia’s small businesses drive massive economic growth, contributing $600 billion to Australia’s economy and employing 5.4 million people. But these businesses face constant cash flow challenges. One in three owners dips into personal savings to pay business bills. One in five operates with no cash reserves. When a customer pays late or equipment breaks, these businesses stand on a knife edge between survival and closure.

This is the challenge that drives Prospa, Australia’s leading digital small business lender. Since 2012, Prospa has lent over $4 billion to small businesses. Prospa’s mission: to unleash the potential of all small businesses by providing fast credit decisions for loans up to $500,000, entirely online, often within two hours.

“Traditional banks systematically underserve small businesses,” explains Jin Foo, Head of Data & Analytics at Prospa. “When a business spots an opportunity to expand or needs to manage seasonal cash flow, waiting weeks for a bank’s decision may mean missing out on that opportunity.”

The stakes are particularly high in Australia, which has one of the highest small business failure rates among developed economies. Half of all small businesses struggle with uneven cash flow. For seasonal businesses—from tourism operators to farmers—one slow season or natural disaster can spell disaster without access to credit. Prospa’s mission is to help unleash the potential of these small businesses with fast, flexible funding.

Eliminating Days of Manual Data Analytics for Loan Applications

When Foo joined Prospa, the company’s rapid growth had created a complex web of disconnected customer data. Business owners often control multiple entities through trusts and shell companies. These complex structures hide relationships between borrowers. A director might appear on one loan application while controlling ten other companies—each with different borrowing histories. Three to four team members spent up to two days every week manually verifying customer identities and relationships across multiple systems.

“The problem with tabular data was that for every scenario, you needed a specific dataset and a specific query,” Foo explains. “You couldn’t have one master view. You’d always miss edge cases because you first had to think of the combination, prepare the data, then write the query. You couldn’t discover things you didn’t already know to look for.”

Relational databases struggle with nested, interconnected data structures like shareholding networks, which require complex SQL joins between tables, leading to poor performance as data complexity grows and more joins are required.

Prospa’s relatively small yet wide-ranging datasets required complex SQL queries across customer relationship management (CRM) data, loan origination records, loan servicing information, government business registries and Credit Bureau data.

Foo recognised that tables were the wrong data structure for mapping complex business relationships. In contrast, Neo4j’s graph database doesn’t require joins at all: it natively models relationships, making it easy to traverse ownership structures and quickly adapt to new requirements. After evaluating various options, including Elasticsearch and other graph databases, Prospa selected Neo4j.

“We chose Neo4j because it’s originally open-source, providing flexibility and standardization,” Foo explains. “If we needed to scale up or change direction, we had options at every stage.”

The team started with a proof of concept on Neo4j’s free Community Edition, loading test data to demonstrate relationship mapping. The results revealed surprising connections between seemingly unrelated businesses—patterns impossible to detect with traditional queries.

Success with the POC led to a staged deployment that included:

1. Initial implementation focusing on customer identity verification;
2. Integration with marketing systems using Segment and Braze;
3. Expansion to full enterprise deployment connecting all customer data; and
4. Development of an intelligence layer powering predictive insights.

While Prospa continues to use its CRM as a master record of customer data, a Neo4j Enterprise graph database now serves as Prospa’s system of record and most trustworthy source of truth for customer relationships and connections across internal systems, integrated with Azure and Databricks for data processing. This creates a unified view across lending, origination, and customer management systems.

Unlocking Greater Agility to Help SMEs Succeed

The impact extends far beyond faster identity verification. “Neo4j has become central to our business,” Foo says. “It’s now the only system that understands who everyone is across all our platforms.”

Most importantly, the solution enables Prospa to better serve small businesses. Prospa is building what Foo calls a decision intelligence layer —combining relationship data with business health indicators to drive smarter lending. The company uses insights from successful customers to create playbooks, helping other businesses thrive. By understanding seasonal patterns and industry-specific challenges, Prospa can better support businesses through difficult periods.

The system analyses transaction patterns and cash flow trends to identify businesses that might need support before they face a crisis.

Key benefits for Prospa’s internal teams include:

• Real-time verification of business relationships and ownership structures
• Automated detection of concerning patterns in loan applications
• Unified customer profiles enabling personalized product offers
• Self-service analytics through natural language queries via Microsoft Copilot

Prospa is testing Microsoft Copilot as a tool to support business users. Prospa employees can ask Copilot complex questions in plain English, for example: “Show me customers at risk of arrears in the next 30 days who we haven’t contacted in 60 days.” Copilot translates this question into Cypher queries against the graph. Cypher is Neo4j’s declarative, schema-flexible graph query language. 

“This allows anyone in our company to access deep insights without knowing how to code,” says Foo. “Sales teams, collections specialists—they can all query the graph directly through natural language. What once took days of data team effort now happens in minutes, letting us focus on helping small businesses succeed.”

Future developments include:

• Adding temporal analysis to track relationship changes
• Expanding natural language interfaces
• Developing predictive models for business health
• Creating industry-specific lending products

Large corporations gain an edge through the sheer scale of data available to them, enabling optimisations across the value chain. Prospa as Australia’s leading digital lender aims to provide the same edge to small businesses through the power of connected data made possible through graphs.

For Australia’s small businesses, this means the difference between survival and success. When a restaurant owner needs to buy equipment for a second location, or a farmer needs to secure inventory before harvest, they now have a financial partner who understands their business and can act quickly. Prospa’s technology does more than approve loans — it helps 5.2 million employees keep their jobs, helps seasonal businesses stay open year-round, and helps small businesses secure a reliable source of capital when opportunities arise.

Get in Touch

Curious about what insights you could unlock for your business with graph-powered solutions? Let’s talk — reach out, and we’ll get in touch.