APA > RPA: A New Standard for Enterprise Automation

For all the hype, most enterprise AI projects are stuck in pilot mode. Nearly 9 in 10 AI applications fail to make it past the proof-of-concept stage, and more than 80% of large companies aren’t seeing a tangible impact on enterprise-level profits from their use of generative AI.

A big part of the problem is trust. Enterprises, particularly those with high regulatory and accuracy requirements like finance, worry about hallucinations and dislike the opacity of black-box systems where they can’t check how the model arrived at its results. They need decisions that are traceable and outcomes that are consistent.

Enter Agentic Process Automation: Combining the reliability of robotic process automation with the adaptability of generative AI. Maisa is solving one of the toughest challenges in AI: making it reliable and safe to use in mission-critical business operations.

This unlocks one of the largest opportunities in enterprise software: finally delivering on AI’s promise at scale. Very few categories offer this combination of horizontal applicability, mission-critical value, and pent-up demand.

CEO David Villalón and CSO Manuel Romero are two exceptional leaders in applied AI. David was previously Chief AI officer at Clibrain (Spain’s answer to Mistral) and Director of Product at Voicemod, and Manuel was formerly Chief Scientific Officer at Clibrain and is one of the top global contributors to Hugging Face. They saw firsthand that existing AI tools couldn’t be trusted for complex, multi-step business processes. So, instead of building another chatbot, they reimagined AI automation from scratch using operating system architecture principles.

Maisa’s platform is built on the Knowledge Processing Unit (KPU), a proprietary reasoning system which uses general-purpose large language models (LLMs). Rather than relying on probabilistic guesswork, Maisa’s digital workers take actions through logical, step-by-step computation. Every action is recorded in a Chain-of-Work: a transparent, auditable record that shows exactly how each outcome was reached.

With Maisa Studio, the company’s agentic process automation platform, anyone in a business without a technical background can create and deploy these digital workers with natural language.  This is powered by a method the company calls HALP (human-augmented LLM processing), where digital workers learn on the job as they complete real work inside an organisation, without heavy upfront training or manual programming. Now you can vibe code your workforce into existence.

We believe Maisa represents the next generation of enterprise AI, and we’re proud to back David, Manuel, and the team as they build real trust, scale, and intelligence into automation.