Coris is building the modern SMB risk & data intelligence platform. Today, we’re excited to introduce Account Graph, which automatically surfaces relationships between existing businesses within a company’s portfolio. 

As the latest addition to our Fuzio product, Account Graph allows risk teams to instantly uncover repeat bad actors. This helps reduce false positives and prevents downstream fraud losses.

‍

Using graph databases to fight SMB fraud 

SMB data is growing exponentially, but software platforms, payment processors and fintechs still lack the tools to extract meaningful insights from this information. This is especially true in SMB fraud: risk analysts still uncover patterns of suspicious behavior through manual analysis – there’s no way to systematically identify and connect these patterns across different accounts.

Graph databases can help surface relationships between SMB data points in a scalable way. Banks and large fintech companies such as PayPal already employ graph databases to identify advanced fraud scenarios within their customer base. However, no one has made this data visualization available to other software platforms and fintechs, who serve the majority of SMBs worldwide.

‍

Introducing Account Graph

Account Graph is the first data visualization tool that enables risk teams to identify patterns within their SMB data in a secure and no-code way:

• Data security: It leverages data specific to each company and doesn’t share data with other partners in Coris’s ecosystem, keeping data private.
• No engineering bandwidth needed: It leverages a customer’s existing Stripe or Adyen data, and automatically uncovers relationships between SMB users at each application submission. This requires zero additional integrations or engineering work for customers.

How does it work?

When a new SMB application is submitted via Stripe or Adyen, Account Graph cross-references this data against the customer’s entire SMB portfolio. It generates signals on how many accounts this application might be linked with based on matches on the following data points:

• Email address, phone, or website (absolute matches)
• Physical address (fuzzy match)

It then categorizes whether these linked accounts are “good” or “bad”. “Good” linked accounts are SMBs who have had a positive history on the platform with few serious risk or fraud signals. “Bad” linked accounts are SMBs who were previously rejected from the platform or were terminated from the platform due to fraudulent activity.

Account Graph converts all of this information into “signals”, which teams can use to set up rules and automated actions within Fuzio. Check out some common use cases in the next section. 

‍

Common use cases

Onboarding

If an incoming SMB application matches information from a “bad” linked account, teams forward the application to manual review or automatically reject it. If the account is forwarded to Case Management, teams can seamlessly review linked account data there. Conversely, a new application with strong links to an existing long-tenured good account can potentially be auto-approved instead of treating it like a brand new account with no history.

Ongoing monitoring

Some savvy fraudulent merchants don’t commit fraud upfront. Instead, they try to develop a positive first impression before engaging in more suspicious behavior. If a merchant “goes bad” or goes out of business, teams can run an Account Graph check to see if there are any other accounts linked to this merchant. They can shut down these linked accounts, preventing future fraudulent activity or losses from the bad actor.

‍

What’s next?

Coris is modernizing the SMB risk & data intelligence process, from data discovery to automated insights and risk decisioning. We’ll continue adding to Account Graph with additional matchable data fields, surfacing 2nd and 3rd degree connections between accounts, and more. 

‍Reach out if you’d like to learn more about Account Graph.