Connect with us

Funding

Maisa Raises $25M to Power Trustworthy, Transparent ‘Digital Workers’

mm
Maisa co-founders: David Villalon (CEO) and Manuel Romera (CSO)

Enterprise AI is plagued by failure—studies show up to 95% of generative AI pilots collapse. But Maisa, bridging Valencia and San Francisco, is engineering accountability into automation. Following a $25 million seed round led by Creandum with backing from Forgepoint Capital, NFX, and Village Global, Maisa is poised to transform AI workflows with agentic process automation that’s as auditable as it is intelligent.

Inside the Engine: KPU & Chain-of-Work — Building Trust from the Ground Up

At the core of Maisa’s platform is the Knowledge Processing Unit (KPU)—a novel architecture that reconceives AI reasoning. The KPU includes:

  • A Reasoning Engine, powered by a large language model, which plans out multi-step workflows.
  • An Execution Engine that carries out these plans and feeds outcomes back for recalibration.
  • A Virtual Context Window that streamlines information flow, focusing the model only on relevant data to mitigate hallucinations.

The genius of the KPU is in how it reshapes the role of the LLM. Instead of acting as a probabilistic text generator, the model is treated as one component inside a disciplined computational framework. The KPU orchestrates reasoning like an operating system, breaking problems into manageable, auditable steps and ensuring each action can be validated. This structure turns a once-opaque model into a predictable system of logic, where errors can be detected, corrected, and prevented from cascading.

Complementing the KPU is the Chain-of-Work—a meticulously logged audit trail tracking every decision, action, and tool involved in a digital worker’s process. Unlike typical AI outputs that leave users guessing, the Chain-of-Work functions like a black box recorder for automation. Every calculation, every data pull, every reasoning step is preserved in detail. This gives enterprises the ability to not only trust outcomes, but to retrace them, replay them, and refine them over time. For compliance-heavy industries, this isn’t just a convenience—it is the foundation of safe deployment at scale.

Maisa Studio in Action: No-Code Agents Rooted in Transparency

Leveraging KPU and Chain-of-Work, Maisa Studio enables “citizen developers”—non-technical staff—to deploy digital workers using plain language instructions. Through Maisa’s HALP (Human-Augmented LLM Processing), the system interacts to clarify intent, constructs workflows, integrates across hundreds of APIs, and begins learning dynamically—all without developers or datasets.

In practice, this has already delivered meaningful impact: a financial services firm cut false positives by 99% and achieved a 10× productivity boost per employee—with full deployment accomplished in just three onboarding sessions. Global banks, automakers, and energy companies are piloting the platform to automate compliance-heavy processes at scale, where transparency and auditability are non-negotiable.

Why It Matters—A Vision for AI You Can Trust and Scale

Maisa doesn’t just offer automation—it delivers accountable AI that embeds trust by architecture. In industries burdened by regulation, opacity, and high stakes—finance, healthcare, energy—the ability to trace every automated decision is paramount. The Chain-of-Work enables businesses to inspect, audit, and validate AI logic at any step, meaning regulators and internal teams need not guess at how conclusions were reached. Instead, they can verify with precision.

Meanwhile, the KPU’s design systematically suppresses hallucinations by isolating reasoning from data noise and structuring execution. This reliability sheds the unpredictability that often makes enterprises wary of AI. Rather than producing outputs shrouded in mystery, Maisa’s agents deliver logical, predictable, and consistent results.

Looking ahead, this platform represents a fundamental shift: AI becomes a trusted collaborator—one whose reasoning is transparent, whose actions are traceable, and whose “thought process” can be refined and audited. As Maisa’s system is model-agnostic, organizations retain flexibility to adopt stronger models in the future—without losing the rigor and oversight of the KPU framework. That adaptability lays the groundwork for sustainable, scalable deployment across evolving enterprise needs.

In essence, Maisa offers a blueprint for AI that isn’t just powerful—but responsible and resilient. In a world where most AI projects fail to deliver, this technology charts a rare path forward—combining innovation with integrity.

Antoine is a visionary leader and founding partner of Unite.AI, driven by an unwavering passion for shaping and promoting the future of AI and robotics. A serial entrepreneur, he believes that AI will be as disruptive to society as electricity, and is often caught raving about the potential of disruptive technologies and AGI.

As a futurist, he is dedicated to exploring how these innovations will shape our world. In addition, he is the founder of Securities.io, a platform focused on investing in cutting-edge technologies that are redefining the future and reshaping entire sectors.