Twin’s first AI agent is an invoice retrieval agent for Qonto customers

Tech Crunch - Mar 27th, 2025

Twin, a Paris-based startup, has unveiled its first AI automation agent, Invoice Operator, in collaboration with Qonto, a fintech company providing business bank accounts across Europe. This new tool aims to streamline the tedious process of invoice retrieval, a task that Qonto's customers previously spent hours on each month. By leveraging OpenAI's CUA model, Twin's agent can automatically fetch, download, and attach invoices to transactions, drastically reducing the need for manual input and bypassing the limitations of traditional automation tools like RPA and API-based systems.

The development of Invoice Operator signifies a significant advancement in AI-driven automation, with Twin showcasing its ability to support thousands of applications within a short time frame. This capability positions Twin as a potential leader in the B2B agentic applications market, with aspirations to expand into other industries such as e-commerce and customer service. As AI agents become more accessible and efficient, Twin's innovation could herald a new era of cost-effective and precise task automation, prompting further integration of AI into everyday business operations.

Fintech Open Ai Ai Automation Twin Qonto Invoice Processing Cua Model B2 B Applications Hugo Mercier

Story submitted by Fairstory

RATING

6.8

Fair Story

Consider it well-founded

The article provides a detailed overview of Twin's innovative approach to automation in partnership with Qonto, highlighting the potential benefits of AI-driven solutions in the fintech sector. It is timely and relevant, capturing a significant development in the industry. However, the article could benefit from more balanced coverage, including potential challenges and ethical considerations associated with AI deployment. While the clarity and readability are strong, the reliance on internal sources and lack of diverse perspectives limit its overall impact and engagement. Addressing these areas would enhance the article's quality and provide a more comprehensive understanding of the topic.

Accuracy

The article presents a generally accurate depiction of Twin's launch and its partnership with Qonto. The claim that Twin emerged from stealth in January 2024 is supported by the context provided, and the description of its Invoice Operator aligns with its stated functionalities. However, some claims, such as the exact number of applications supported by Twin and the specific use of OpenAI's CUA model, require verification to ensure precision. The story accurately captures the essence of Twin's technological advancements and its application in the fintech space, but it would benefit from corroborating details about the team size and the exact scale of operations.

Balance

The article primarily focuses on Twin's technology and its partnership with Qonto, offering a positive perspective on the innovation's potential. While it effectively highlights the benefits of AI-driven automation, it lacks a balanced view by omitting potential challenges or criticisms of AI agent deployment. The absence of perspectives from external experts or competitors in the automation sector limits the article's balance. Including insights into potential drawbacks or industry challenges would provide a more rounded view.

Clarity

The article is well-written, with a clear structure and logical flow that makes it easy to follow. The language is straightforward, and the technical aspects of Twin's product are explained in a manner accessible to a general audience. The use of specific examples, such as the process of invoice retrieval, enhances understanding. However, some technical jargon related to AI and automation could be further simplified for readers unfamiliar with the field.

Source quality

The article seems to rely heavily on statements from Twin's co-founder and CEO, Hugo Mercier, which may introduce some bias. While these are credible sources for information about the company's operations, the lack of external sources or independent verification of claims weakens the overall source quality. The story would benefit from including data or opinions from industry analysts or third-party experts to enhance credibility and provide a more comprehensive understanding of the technology's impact.

Transparency

The article provides a clear overview of Twin's product and its intended use case but lacks transparency in terms of methodology and potential conflicts of interest. There is no disclosure of how the information was gathered or any potential biases from the sources. Additionally, the article does not address possible limitations or ethical considerations of AI deployment in financial services, which would contribute to a more transparent narrative.