Gabriel Muñoz of VASS Intelygenz talks Llama 4 models for open-source AI with VKTR

Gabriel Muñoz of VASS Intelygenz talks Llama 4 models for open-source AI with VKTR

Gabriel Muñoz shared his expert insights with Scott Clark on Meta’s Llama 4, highlighting its improved efficiencies, enhanced safety measures, and how it compares to other open-source models, while also exploring what lies ahead for the industry. Speaking on practical implementation, he noted that at VASS Intelygenz, “model evaluation is aligned with clearly defined business objectives and acceptance criteria. We apply rigorous real-world validation through shadow testing, combining automated metrics with human review to ensure our models meet both technical and business standards.”