Blab AI
Rivian Uses Amazon Bedrock AI to Cut 15-Day Manual Work in Finance Close
Electric‑vehicle maker Rivian has deployed an AI system built on Amazon Bedrock to automate the accrual of purchase orders for custom manufacturing tools. According to an AWS blog post dated August 13 2026, the new system eliminates more than 15 days of manual work per month‑end close cycle.
The accounting problem is complex. Custom tooling—such as stamping dies and injection molds—can take one to two years to produce, yet Rivian often does not receive an invoice until 18 months or more after the order is placed. Accounting rules require the cost to be recognized gradually over the life of the tool, not all at once when the invoice arrives. Rivian’s finance team needed a way to track the amount that should be accrued each month.
Instead of hard‑coding the rules into software, Rivian stored its actual accounting procedures as plain‑text instructions that the AI reads directly. A finance manager can update the process by editing a document, and the AI picks up the change automatically the next time it runs. The system monitors new purchase orders that need attention, consults the instructions, emails the appropriate person if information is missing, calculates the accrual amount, and drafts an accounting entry. A human still signs off on every entry before it is posted to the books.
Yogesh Yadav, senior director of global finance systems at Rivian, said the partnership with AWS and the implementation of AI‑powered automation “represents a fundamental shift in how we approach finance operations.” He added that the new approach has eliminated weeks of manual work each month and created a scalable foundation that will support the company’s growth.
The human approval step is intentional. It satisfies Rivian’s internal controls and its external auditors, who found the new process easier to trace than the spreadsheets it replaced. When a manager corrects an error that the AI made, the correction is fed back into the written instructions, so the same mistake should not recur.
Rivian built a proof‑of‑concept prototype in five weeks before putting the system into production. The company is already planning to apply the same approach to other finance tasks, such as working‑capital monitoring and cash‑flow tracking.
The move fits a broader trend. A PYMNTS Intelligence report released in December 2025 found that 45 % of chief financial officers reported using AI in finance today, while another December report noted that 66 % of accounts‑payable teams had experienced an increase in manual workload over the previous year.
Rivian’s solution is built on Amazon Bedrock, a cloud service that provides a unified API to access foundation models from multiple AI vendors. Bedrock is designed for enterprise use and competes with similar platforms from Microsoft and Google.
The AI agents that Rivian uses are part of Bedrock’s AgentCore framework. The agents read the plain‑text instructions, interact with the company’s SAP finance system, and generate draft entries that are then reviewed by finance staff.
The system’s impact on Rivian’s month‑end close is significant. By automating the accrual of custom tooling, the company saves more than 15 days of manual work each cycle, which translates into weeks of labor saved each month.
The approach also improves auditability. Because the AI follows explicit, human‑editable instructions, auditors can trace the logic behind each entry more easily than with opaque spreadsheet models.
Rivian’s experience illustrates how generative AI can be applied to highly regulated, rules‑based finance functions without sacrificing control or compliance. The company’s next steps include expanding the AI‑driven automation to other finance processes and exploring additional use cases within its broader corporate finance operations.
In summary, Rivian’s partnership with AWS and the use of Amazon Bedrock AI agents have streamlined a complex accounting task, reduced manual effort, and strengthened audit trails. The company plans to extend this approach to other finance functions, positioning it as a potential model for AI adoption in corporate finance.
The accounting problem is complex. Custom tooling—such as stamping dies and injection molds—can take one to two years to produce, yet Rivian often does not receive an invoice until 18 months or more after the order is placed. Accounting rules require the cost to be recognized gradually over the life of the tool, not all at once when the invoice arrives. Rivian’s finance team needed a way to track the amount that should be accrued each month.
Instead of hard‑coding the rules into software, Rivian stored its actual accounting procedures as plain‑text instructions that the AI reads directly. A finance manager can update the process by editing a document, and the AI picks up the change automatically the next time it runs. The system monitors new purchase orders that need attention, consults the instructions, emails the appropriate person if information is missing, calculates the accrual amount, and drafts an accounting entry. A human still signs off on every entry before it is posted to the books.
Yogesh Yadav, senior director of global finance systems at Rivian, said the partnership with AWS and the implementation of AI‑powered automation “represents a fundamental shift in how we approach finance operations.” He added that the new approach has eliminated weeks of manual work each month and created a scalable foundation that will support the company’s growth.
The human approval step is intentional. It satisfies Rivian’s internal controls and its external auditors, who found the new process easier to trace than the spreadsheets it replaced. When a manager corrects an error that the AI made, the correction is fed back into the written instructions, so the same mistake should not recur.
Rivian built a proof‑of‑concept prototype in five weeks before putting the system into production. The company is already planning to apply the same approach to other finance tasks, such as working‑capital monitoring and cash‑flow tracking.
The move fits a broader trend. A PYMNTS Intelligence report released in December 2025 found that 45 % of chief financial officers reported using AI in finance today, while another December report noted that 66 % of accounts‑payable teams had experienced an increase in manual workload over the previous year.
Rivian’s solution is built on Amazon Bedrock, a cloud service that provides a unified API to access foundation models from multiple AI vendors. Bedrock is designed for enterprise use and competes with similar platforms from Microsoft and Google.
The AI agents that Rivian uses are part of Bedrock’s AgentCore framework. The agents read the plain‑text instructions, interact with the company’s SAP finance system, and generate draft entries that are then reviewed by finance staff.
The system’s impact on Rivian’s month‑end close is significant. By automating the accrual of custom tooling, the company saves more than 15 days of manual work each cycle, which translates into weeks of labor saved each month.
The approach also improves auditability. Because the AI follows explicit, human‑editable instructions, auditors can trace the logic behind each entry more easily than with opaque spreadsheet models.
Rivian’s experience illustrates how generative AI can be applied to highly regulated, rules‑based finance functions without sacrificing control or compliance. The company’s next steps include expanding the AI‑driven automation to other finance processes and exploring additional use cases within its broader corporate finance operations.
In summary, Rivian’s partnership with AWS and the use of Amazon Bedrock AI agents have streamlined a complex accounting task, reduced manual effort, and strengthened audit trails. The company plans to extend this approach to other finance functions, positioning it as a potential model for AI adoption in corporate finance.
