How Auditoria.ai Brings Trusted Autonomous AI Agents to Finance Teams

Rohit Gupta of Auditoria.ai (pictured right)

The spring event season was dominated by discussions of autonomous agents, but clarity on their practical implementation remained elusive. While vendors initially hyped full autonomy, many later emphasized human-in-the-loop approaches to address customer skepticism.

Workday's AI Partner Ecosystem

At Workday's Innovation Summit, Auditoria.ai stood out with its purpose-built agentic AI solutions for CFOs. The company is part of Workday's new Agent Partner Network, which includes major players like AWS, Adobe, and Accenture.

Auditoria.ai CEO Rohit Gupta explained their unique approach:

We allow users to run our software in collaborative (human-in-the-loop) mode initially, then transition to full autonomy as trust builds.

Building Trust Through Gradual Autonomy

Key aspects of Auditoria.ai's strategy:

• Finance-specific language models trained on accounting/procurement data
• 300+ language/currency support for global operations
• 10-12 week trust-building period before enabling autonomous mode
• Granular KPIs measuring time savings, cost reductions, and risk mitigation

Gupta emphasized that finance teams demand near-perfect accuracy, not the 70-80% acceptable in other domains. Their system reportedly achieves 90%+ accuracy on financial queries and high 90s on document processing.

Measuring Success

Auditoria.ai tracks:

1. Efficiency gains (time/cost savings)
2. Spend anomaly detection
3. Risk mitigation (duplicate invoices, fraud prevention)

The Road Ahead

Auditoria.ai is piloting a strategic finance analyst agent that combines Workday data with feeds from Bloomberg/Capital IQ. Planned for launch at Workday Rising, this could further transform financial operations.

For more on Workday's AI developments, see Phil Wainewright's coverage.