Qonto, the financial management platform for European SMEs, has launched Invoice Operator, an AI-driven solution to streamline invoice collection. Developed with Twin and powered by OpenAI’s Conversational User Agent (CUA) model, it eliminates the hassle of manual invoice retrieval.
Simplifying invoice collection
Qonto users spend over three hours per month tracking invoices from multiple platforms. Across Qonto’s 500,000 customers, this adds up to 1.2 million invoices monthly. Previous automation efforts struggled due to a lack of APIs and fragile RPA scripts that frequently broke. Invoice Operator changes the game.
How it works
Users log into twin.so/qonto, and the AI agent takes over:
• Identifies missing invoices from bank transactions
• Autonomously retrieves invoices from provider websites
• Matches and uploads invoices to Qonto
Unlike traditional RPA, Twin Agent adapts to different interfaces, offering human-like flexibility with API-level reliability. In beta testing, 1,500 users retrieved 500 invoices from 50 providers in under 10 minutes.
A new standard in AI automation
Reliability
• Learns from interactions and self-improves
• Manages mismatched amounts, currencies, and dates
Security
• Destroys credentials after sessions
• Implements compliance with GDPR and SOC 2 Type 2
Scalability
• Operates in a custom-built Chromium environment
• Maintains sub-100ms latency at high volumes
Qonto’s Invoice Operator redefines financial automation, making invoice collection faster, smarter, and more secure.
