A Chat with David Villalón, Co-Founder & CEO at Agentic Process Automation Platform: Maisa

Tell us about Maisa

 

Maisa is an enterprise Agentic Process Automation (APA) platform that creates and manages AI-powered Digital Workers to automate complex, knowledge-intensive processes with full traceability. Unlike traditional AI agentic solutions, our ‘Chain-of-Work’ concept ensures compliance and transparency by combining deterministic, code-based execution with auditable explainability.

Our Maisa Studio enables users to configure and deploy accountable AI Agents using natural language, integrating them with enterprise systems while leveraging a basically Infinite Virtual memory (Virtual Context Window) to store and retrieve large volumes of structured knowledge and know-how.

We solve scalability, maintainability, and operational cost challenges by consolidating credential management, data indexing, tool integrations, and continuous improvement into a single platform. Our Digital Workers automate workflows while capturing and refining organisational know-how, continuously improving through real-time feedback.

By addressing key limitations in existing process automation solutions (RPA), we provide enterprises with a robust, scalable, and future-proof approach to intelligent process automation.

 

 

How did you come up with the idea for the company?

 

Maisa was born from our first-hand experience with AI models and RAG-based approaches, where we saw their flaws that prohibit their use in business automation at scale. Despite the excitement around AI, the probabilistic nature of LLMs—especially with RAGs—led to compounding error rates in multi-step processes, making them unreliable for enterprise use. As token outputs increase, so do risks, proving existing AI methods lacked deterministic reliability.

We went back to the drawing board, drawing inspiration from operating system architecture models to build a fundamentally different AI approach. This led to Maisa and the Knowledge Processing Unit (KPU), designed for full transparency, traceability, and deterministic execution. Our Chain-of-Work methodology ensures every decision is provable, making it the first AI system capable of automating complex, high-stakes business processes with complete auditability.

 

 

What is your advice to aspiring entrepreneurs?

 

Learn to see around corners. Great founders don’t just react to the present; they anticipate shifts in technology, markets, and human behavior.

Stay curious and be a fast learner. With 1,000 hours of deep focus, you can master anything. Young, open minds have an advantage because they aren’t constrained by outdated paradigms—use that to your benefit.

Finally, dare to be contrarian. If you’ve identified a problem and believe in a solution others dismiss, be brave and stick to your convictions. Challenge conventional wisdom, test your ideas, and be open to being wrong—but don’t fear thinking differently. The biggest breakthroughs come from pushing against the status quo.

 

What most excites you about Maisa?

 

Our AI orchestrated system employs a computational approach to reasoning, executing each step in a transparent, deterministic manner. It leverages internal capabilities while interacting with external tools, ensuring precision and adaptability.

By remaining tool-agnostic and embracing this interactive methodology, we unlock limitless possibilities for enhancing AI quality.

 

How has the company evolved over the last couple of years?

 

We started with just my co-founder and a small team, building a proof of concept that showcased our technology’s potential. That early work won the trust of top investors, raising $5M to accelerate our vision.

With funding in place, we grew the team, developed our first product version, and tested it rigorously. Through continuous iteration, we launched with enterprise customers, refined our approach, and now, with strong product-market fit, we’re scaling our go-to-market strategy.

 

What can we hope to see from Maisa in the future?

 

We’re already enhancing supply chain reliability, ensuring compliance, and improving decision-making for major global businesses.

Looking ahead, we’ll expand into industries requiring reliable, auditable process automation, including financial services, energy, manufacturing, and cybersecurity.

A key driver of this evolution is our AI Agents’ continuous self-learning, stored in virtually infinite memory. The system captures know-how from documents and client interviews, allows managers to inject real-world expertise for continuous improvement, and leverages scalable memory via our Virtual Context Window for efficient data processing.

These capabilities make our AI solutions smarter, more efficient, and increasingly adaptive over time.