Maisa AI Secures $25M Funding to Combat 95% Failure Rate in Enterprise AI Solutions

A staggering 95% of generative AI pilots at companies are failing, according to a recent report published by MIT’s NANDA initiative. But rather than giving up on the technology altogether, the most advanced organizations are experimenting with agentic AI systems that can learn and be supervised.

Enter Maisa AI, a forward-thinking startup that has emerged with a promising approach to enterprise automation. With a recent $25 million seed funding round led by European VC firm Creandum, Maisa is transforming the landscape of AI in business through its innovative platform, Maisa Studio. This model-agnostic self-serve platform enables users to deploy digital workers that can be trained using natural language, prioritizing accountability over traditional opaque AI systems.

While some may see Maisa’s model as similar to other so-called vibe coding platforms like Cursor or Lovable, CEO David Villalón emphasizes a key distinction. “Instead of using AI to build the responses, we use AI to build the process that needs to be executed to get to the response — what we call ‘chain-of-work,’” he explained to TechCrunch. This focus on processes creates a more structured approach to AI that businesses can rely on.

The brilliance behind this innovative process is largely attributed to co-founder and Chief Scientific Officer Manuel Romero, who collaborated with Villalón at the AI startup Clibrain. Their partnership formed a response to the critical challenges of AI reliance, particularly the phenomenon of hallucinations in generative models. Villalón noted that their concern stemmed from the impossibility of humans thoroughly reviewing extensive reports in just a few minutes of work.

To address the complex demands of enterprise solutions, Maisa has developed HALP (Human-Augmented LLM Processing). This unique method engages users by asking questions about their needs while its digital workers delineate each procedural step that must be followed. This approach fosters a more interactive and precise environment for responsible AI deployment.

Maisa has also pioneered the Knowledge Processing Unit (KPU), a deterministic system that addresses the real issue of hallucinations found in generative AI systems. From its inception, Maisa has focused on trustworthiness and accountability, resonating with enterprises that require reliable AI for mission-critical tasks. High-profile clients in sectors such as banking, car manufacturing, and energy already benefit from Maisa’s cutting-edge technology.

This enterprise-first mindset positions Maisa as a more sophisticated alternative to traditional robotic process automation (RPA) systems. Unlike standard RPA solutions that rely on rigid, predefined rules and extensive manual programming, Maisa provides a flexible framework that boosts productivity while minimizing the administrative burden on companies. Customers can choose deployment via Maisa’s secure cloud or opt for on-premise solutions, catering to diverse operational preferences.

Although Maisa’s customer base is relatively small compared to popular freemium vibe-coding platforms, the startup is strategically positioning itself to capture the enterprise market as these platforms seek to broaden their reach. Maisa Studio is designed to facilitate this growth, offering a seamless experience for new clients while addressing the complexities of AI implementation.

To capitalize on its promising start, Maisa plans to expand its operations in alignment with existing customers who have a multi-national footprint. With headquarters in both Valencia and San Francisco, the startup is well-positioned in the U.S. market, evidenced by its earlier $5 million pre-seed round, led by esteemed venture firms NFX and Village Global. Recent participation from U.S.-based Forgepoint Capital International, through a partnership with Spanish bank Banco Santander, highlights Maisa’s appeal within regulated sectors, enhancing its credibility.

Focusing on complex use cases that demand accountability and transparency from non-technical users could be a major differentiator for Maisa. As Villalón points out, this “AI framework gold rush” presents opportunities as well as challenges, warning that what seems like a “quick start” may turn into a troublesome journey when reliability, auditability, and error correction become necessary.

Aiming to scale effectively, Maisa is motivated to use its funding strategically to double its workforce from 35 to approximately 65 employees by early 2026. With a waiting list of eager clients, the startup anticipates rapid growth in the coming months. “We are going to show the market that there is a company that is delivering what has been promised, and that it’s working,” Villalón confidently stated.