Our AI Research

We are improving the way our AI models access, select, and use knowledge. Our goal is to increase the reliability and relevance of responses by optimizing every step of the retrieval-augmented generation process. We are exploring new strategies for retrieval, orchestration, and reasoning to create increasingly efficient AI agents.

We develop AI agents capable of planning, reasoning, and acting autonomously in complex environments. These systems learn to make decisions, coordinate actions, and adapt to new contexts, while remaining aligned with human and business needs. Our goal is to design truly collaborative AI systems that enhance human capabilities rather than replace them.

We design reliable, secure, and resource-efficient language models and AI systems. Our approach combines the creation of high-performance models, architectural optimization, and control over their energy footprint. We are working to enhance the robustness, security, and transparency of the generated responses, while developing language models that are more scalable, resource-efficient, and ethically responsible. In this way, we are envisioning the AI systems of tomorrow—capable of automating complex tasks with scientific rigor, while reducing their environmental impact.