Revolutionizing Invoicing: Twin Launches AI Agent for Qonto Customers
In January 2024, Twin emerged from stealth mode, introducing a groundbreaking automation solution that leverages artificial intelligence. The Paris-based company has partnered with Qonto, a prominent fintech startup serving over 500,000 business account customers across Europe, to launch its innovative automation agent.
The Rise of AI Automation Agents
As businesses look for ways to automate repetitive tasks, several solutions are available. Companies often rely on API-based, no-code, or low-code automation products like Zapier, while others may opt for Robotic Process Automation (RPA) software such as UiPath. However, Twin believes there is a more effective approach to automation.
Introducing Invoice Operator
Twin’s first product, Invoice Operator, is designed specifically for Qonto and illustrates the advantages of utilizing artificial intelligence in business operations. Qonto processes millions of invoices monthly, and customers typically spend countless hours collecting and uploading these documents.
Over a three-month development period, Twin has created an automatic invoice-retrieval tool that streamlines this process significantly. Here’s how it works:
- When users launch Invoice Operator, it retrieves a list of transactions that have missing invoices.
- The tool displays the necessary services to access for downloading invoices next to a browser window showing the agent’s actions.
- If user login credentials are required, the browser pauses for manual input before continuing the automated process.
- Once logged in, Twin’s Invoice Operator locates past transactions, downloads invoices, and automatically attaches the PDFs to the respective transactions in the Qonto account.
Why Choose AI Over Traditional Automation?
Twin’s co-founder and CEO, Hugo Mercier, highlighted the challenges of traditional automation methods during a product demonstration. He explained that handling automation at Qonto’s scale requires addressing a vast array of services—potentially hundreds of thousands—which would be unmanageable with standard RPA. This approach necessitates creating custom scripts for each website, requiring continuous updates with any changes.
In contrast, API-based solutions like Zapier take significant time to grow their application support. For instance, it took Zapier a decade to integrate 8,000 applications, whereas Twin has already achieved support for thousands of applications within months of launching Invoice Operator.
Powered by Advanced Technology
At its core, Twin operates a Chromium-based web browser on a server and utilizes OpenAI’s CUA (computer-using agent) model. Notably, Twin was among the select 15 companies to beta test this innovative technology, which also powers OpenAI’s Operator, a product designed for prosumers to automate tasks.
With a focus on user-friendliness, Twin aims to simplify the experience for end-users, particularly those who are not tech-savvy. Users simply log into their accounts and launch the tool, which then navigates to retrieve invoices without requiring complex configurations.
Future Prospects for AI Agents
Looking ahead, Twin envisions a world in which AI agents can enhance various industries by automating tasks such as:
- Managing orders for e-commerce businesses
- Classifying marketplace catalogs
- Retrieving information for call center agents
Twin is advocating for a future where AI agents not only become more affordable and efficient but also more accurate across diverse tasks. As the company continues to refine the core agent platform that powers Invoice Operator, it remains to be seen how developers can leverage this technology for their applications.
For further insights on automation and fintech innovations, visit Qonto’s official website or explore more about Twin’s offerings.
