Mario Garrido

System architect and AI infrastructure engineer. I define architecture for agent systems and foundation model infrastructure. The agent execution architecture I defined is the required foundation for all agent initiatives at Kuona; the forecasting platform architecture I redesigned underpins the company's foundation model development.

Architectural Impact

• All agent initiatives at Kuona operate through the ExpertAI system architecture I defined — org-wide execution, orchestration, and evaluation patterns.
• Foundation model development follows the forecasting platform system decomposition I defined; the data science team has followed this architecture for approximately two years.
• Forecasting infrastructure redesign reduced onboarding new client data from days to hours and is enabling real-time forecasting capabilities.
• Deprecated per-client training-on-demand systems (2022–2024) and replaced them with a unified generalized platform spanning 500+ servers — onboarding new configurations went from weeks to days; improved forecasting accuracy through standardized infrastructure.
• Provide architectural guidance across teams, review and approve system designs, and define infrastructure standards across engineering and data science.