A shocking of 95% generative AI pilots in companies FailingAccording to one recently Report Published by Nanda Pahal of MIT. But instead of completely leaving technology, the most advanced organizations are experimenting with agentic AI systems that can learn and look after.
Where is he Mysa Ai Comes in. The year -old startup has created its full approach around the base that an AI agents are required for enterprise automation, not opaque black boxes. With a new, $ 25 million seed round led by European VC firm Creandum, it is now launched Massa studioA model-unknowable self-service platform that helps users to deploy digital workers who can be trained with natural language.
While it may seem familiar-the so-called vibe coding such as cursor and crand-supported cute is reminiscent of platforms-Mesa argues that its approach is fundamentally different. “Instead of using AI to create reactions, we use AI to create the process that we need to execute-which we say ‘chain-off-work’,” Massa CEO David Wilon told Tekkachch.
The process behind this process is the co-founder and chief scientific officer of Massa, manual Romero, who previously worked with Wilalon in the Spanish AI Startup Clibraine. In 2024, the pair formed a team to create a solution to hallucinations after seeing this for the first time that “you can’t trust AI,” Willalon said.
The pair is not doubtful about AI, but they feel that it will not be possible for humans to review “three months’ work done in five minutes”. To address it, appoints a system called Maisa Halp, which stands for human-firm LLM processing. This custom method acts like students on the blackboard – it asks users about their needs while digital workers underline each step that they will follow.
Startup also developed the knowledge processing unit (Kpu),, A determinable system designed to limit halvery. While MAISA started with this technical challenge rather than an use case, it was soon found that its stakes on reliableness and accountability resonated with companies that were expecting AI to implement AI in important tasks. For example, customers using Maisa currently in production include a large bank, as well as companies in car manufacturing and energy sectors.
By serving these enterprise customers, Maisa expects to bring itself into a position as a more advanced form of robotic process automation (RPA) that unlocks productivity gains without the need of companies to rely on strict predetermined rules or comprehensive manual programming to companies. To meet their requirements, the startup provides them deployment to either their safe clouds or through on-arrogance sins.
Techcrunch event
San francisco
,
27-29 October, 2025
This enterprise-first approach means that the customer base of Maisa is still much lower than millions of people for the Frimium Vibi-Coding platforms. But as these platforms are now looking for enterprise customers, moving in the opposite direction with Massa Maisa Studio, which is designed to enhance and adopt its customer funnel.
Startup plans to expand with existing customers who have operations in many countries. With the double headquarters in Valencia and San Francisco, Maisa already has a leg in the US, as reflected in its cap table; Its $ 5 million pre-east round Last December was led by San Francisco -based venture firm NFX and Village Global.
In addition, TechCrunch specifically learned that American firm Forgepoint Capital International participated in this new era, through it European Joint Enterprise Spanish bank with banks centenn, highlighted its appeal for regulated areas.
Focusing on matters of complex use that seek accountability from non-technical users can be a discrimination for Maisa, whose contestants include Crewai and many other AI-operated, business-focused workflow automation products. In a LinkedIn Post, Villalon Thrown light on This “AI Framework Gold Rush,” warns that “Quick Start” becomes a long nightmare when you need reliability, auditability, or the ability to fix what went wrong. ,
Doubling its goal to help AI scale, Massa planned to use its funding to increase the demand from 35 to more than 35 people by the first quarter of 2026 to fulfill the demand. Starting in the last quarter of this year, the startup estimates rapid development as it begins to serve its waiting list. “We are going to show the market that there is a company that has been promised, and it is working,” said Wilalon.
