Dura Software cuts finance processing time by more than 90% with AI

San Antonio-based Dura Software, which acquires and operates niche business-to-business software companies, is using artificial intelligence to take repetitive preparation work off its finance team’s plate without removing human review from the process.

Dura Software connected Claude to NetSuite through the NetSuite AI Connector Service. The setup allows Claude to retrieve financial data within the limitations of given access roles and permissions, follow reusable instructions, and prepare recurring work for an accountant to review.

“At this point, most of our team members feel the AI Connector Service has enabled Claude to become another team member they can work with to help compile information, improve analysis, and accelerate decision-making,” said Dura Software CFO, Sloan Session.

According to Session, Dura Software’s early use cases focus on work with clear inputs and repeatable steps. Claude can prepare bad-debt and cash-sweep journal entries, assemble reports for the CEO and business operators, and support financial analysis that previously required employees to build saved searches.

Preparation rather than final decision-making is a central theme of Dura Software’s deployment. Finance staff review the output, resolve exceptions, and retain responsibility for controls and approvals before anything is posted or distributed.

“CFOs don’t want AI to bypass finance controls; they want it to remove manual work while operating inside those controls,” said Craig Sullivan, Group Vice President, Product Management at Oracle NetSuite. “With NetSuite AI Connector Service, companies like Dura Software can provide popular AI models with access to their governed system of record. Based on their needs, AI can review the data, prepare the reports and analysis, and assist with business workflows before handing work back to the finance team for review.”

The connector uses Model Context Protocol (MCP), an open technical standard that lets AI systems interact with business applications. NetSuite says the service enables a user to apply its existing role-based permissions and governance policies to configure the connector, limiting the AI model’s access to only the information relevant for the employee’s assigned role.

Configuring AI access controls consistently with NetSuite’s existing identity, permission, and audit framework allows companies to expand automation without creating a parallel control environment, said Sullivan.

“AI should not become a side door into the ERP. It should operate with the same permissions, approval rules, and accountability as the person using it. That is how finance teams can automate more work without giving up control.”

According to Sullivan, the ability to connect popular AI models with NetSuite, in addition to providing embedded AI capabilities, is important as finance teams are unlikely to use a single AI interface for every task.

“An analyst may want an open-ended way to explore performance, while an accountant completing a monthly close task may need a tightly defined workflow with consistent outputs. The goal is to allow organizations to connect secure business information to the workflows and systems that make sense for them while at the same time enabling them to continue to benefit from the extensive AI capabilities built directly into NetSuite,” said Sullivan.

Dura Software is taking an incremental approach to AI adoption. Session recommends giving employees a few practical examples, developing internal champions, and adding one workflow at a time.

“If you take it week by week and try to automate one workflow, build one skill file, or create one scheduled agent, you’ll look back six months later and your organization will look completely different,” Session said.

This approach gives Dura Software a way to test output and controls before expanding access. It also helps to create a clearer ROI test: comparing the time required for a journal entry, report, or analysis before and after automation.

For finance leaders, Dura Software’s experience suggests that the near-term value of AI lies in reducing the manual work surrounding established processes. AI can handle preparation tasks, and finance employees retain judgment and accountability.

That equation may shift as NetSuite is rolling out NetSuite Next, which embeds Ask Oracle as it’s foundational AI capability, enabling conversational intelligence, richer contextual experiences and agentic workflows at the center of its next-generation ERP system. As those native capabilities expand, finance teams may handle more analysis and automation inside NetSuite, while the AI Connector Service gives them access to third-party models when connection to additional systems is needed.

The technology is moving too quickly for CFOs to bet on a single model or delivery method; working with a vendor that supports both embedded AI and controlled external connections can help finance teams adopt new capabilities and remain at the leading edge while maintaining control of access to their data.