Rivian has implemented an AI-based system built on Amazon Bedrock to automate the tracking of accruals for custom manufacturing tools. This initiative has reduced manual workload by over 15 days per financial close cycle. The company previously struggled with accounting for tooling, such as stamping dies, which can take up to two years to produce and often involve delayed invoicing.
Instead of using rigid hard-coded rules, the system utilizes plain-written instructions that finance managers can update by editing a document. The AI monitors purchase orders, calculates amounts, and drafts accounting entries. A human supervisor must approve every entry before it is finalized, ensuring compliance with internal controls and external audit requirements.
The system incorporates a feedback loop where manual corrections made by managers are integrated back into the instructions to prevent recurring errors. Rivian developed the proof of concept in five weeks. Following this success, the company plans to apply this automation model to other financial operations to support its ongoing growth and scalability.
According to a December report by PYMNTS Intelligence, 45% of CFOs currently utilize AI for structured finance tasks like cash flow tracking. Additionally, 66% of accounts payable teams reported an increase in manual workloads over the past year, highlighting the industry-wide demand for the type of automation Rivian has adopted to streamline its financial processes.
