An amazing 95% of the generative pilots of AI in companies They are failingAccording to a recent one report Posted by the Nanda del Mit initiative. But instead of giving up the technology completely, the most advanced organizations are experiencing with the agent systems that they can learn and supervise.
That’s where Maisa Ai Enter. The one -year startup has built its full approach around the premise that business automation requires responsible for officials, not in opaque black boxes. With a new round of $ 25 million seeds directed by the European firm of VC Creandum, it has now launched Maisa StudioAn agnostic self -service platform that helps users to implement digital workers who can be trained with natural language.
While that may sound familiar, reminiscent of the so -called environments coding platforms such as cursor and the adorable backed by Creandum, Maisa argues that his approach is fundamentally different. «Instead of using AI to build the answers, we use AI to build the process that must be executed to get to the answer, what we call ‘work chain,» said the CEO of Maisa, David Villalón, to TechCrunch.
The main architect behind this process is the co -founder and scientific director of Maisa, Manuel Romero, who had previously worked with Villalón at the Spanish Startup Clibrain. In 2024, the duo joined to build a solution to hallucinations after seeing first -hand that «you couldn’t trust AI,» Villalón said.
The couple is not skeptical about AI, but they think it will not be feasible for humans to review «three months of work done in five minutes.» To address this, Maisa uses a system called Halp, which represents the processing of LLM accommodated by humans. This personalized method functions as students in Blackboard: asks users about their needs, while digital workers describe every step they will follow.
The beginning also developed the knowledge processing unit (Kpu), a deterministic system designed to limit hallucinations. While Maisa began from this technical challenge instead of a case of use, she soon discovered that her commitment to reliability and responsibility resonated with companies that expected AI apply to critical tasks. For example, customers who currently use Maisa in production include a large bank, as well as companies in car manufacturing and energy sectors.
By serving these business clients, Maisa hopes to position themselves as a more advanced form of automation of robotic processes (RPA) that unlocks productivity profits without demanding that companies trust in rigid predefined rules or extensive manual programming. To meet their needs, the startup also offers them implementation in their safe cloud or through local implementation.
Techcrunch event
San Francisco
|
October 27, 2025
This business approach means that Maisa’s customer base remains very small compared to the millions of Freemium environments coding platforms. But since these platforms are now exploring how to win business clients, Maisa is moving in the opposite direction with Maisa Studio, which is designed to grow her client fun and relieve adoption.
The startup also plans to expand with existing customers who have operations in multiple countries. With the dual headquarters in Valencia and San Francisco, Maisa already has a point of support in the USA., As reflected in her lid table; is $ 5 million in the seed round Last December was led by risk companies based in San Francisco NFX and Village Global.
In addition, Techcrunch learned exclusively that the American firm ForgePoint Capital International participated in this new round through its European joint company With the Spanish Bank Banco Santander, highlighting its attraction for the regulated sectors.
Focusing in complex use cases that require the responsibility of non -technical users could be a DRAFT of Maisa, whose competitors include Crewai and many other workflow automation products focused on the business with AI. In a LinkedIn post, Villalón outstanding This «AI Framework Gold Rush», warning that «fast start» becomes a long nightmare when it needs reliability, auditability or the ability to fix what went wrong. «
Duplicating its goal to help the AI, Maisa plans to use their financing to grow from 35 to 65 people for the first quarter of 2026 to meet demand. From the last quarter of this year, the startup anticipates rapid growth as its waiting list begins to serve. «We are going to show the market that there is a company that is delivering what has been promised and that is working,» Villalón said.
