Clous Product

Built with my co-founder, Álvaro Villalba, Clous is an AI-native system for HR teams. Designed to help them stop being busy so they can evolve from reactive to proactive.

The agents automate time-wasting tasks, capture better data, and connect the dots so HR teams can focus on doing more for their talent.

My role as product leader was to shape that vision into a practical solution—from market research to UX/UI design, prompt engineering, and designing agentic architectures.

AI-native System

Building agents requires creativity and flexibility, especially when each component relates dynamically to the rest. We ended up with over 100 AI models and quickly realized it wasn't sustainable. We unified them, worked with variables, and made the system manageable through constant iteration—combining evaluations (Evals) and manual review.

Agentic Experience

We didn't want HR teams navigating between pages, wasting time or having to learn where and how to do everything. We wanted a unified interface, and AI was the enabler.

The challenge was designing a dynamic interface that could adapt the UX to each user's needs—where they only had to write what they needed as if asking a colleague.

We started with chat as a liquid interface, but quickly moved beyond it. We used various visual and interactive components that allow the system to reshape itself based on context, making the experience feel natural rather than constrained by traditional UI patterns.

Beyond Chat and Search

Moving beyond chat wasn't just a design strategy—it was about functionality. Our vision was to build the talent data infrastructure that works with high autonomy, capable of absorbing almost any HR user query.

External Systems

Meet users where the work happens. Having access to where work takes place is essential to capture talent data and ensure adoption across all stakeholders.

Operating Model

Not just retrieving context, but organically creating it with all the data across talent management and business, so every response is highly targeted.

Multi-step Tasks

Agent UX means autonomy—letting AI complete end-to-end workflows. This includes scheduled tasks that run in the background, automating repetitive work so teams can focus on what matters.