Daniel Mulet is a Partner at Radical Ventures who has personally led the firm's thesis work across multiple technically demanding areas—climate, life sciences, and advanced materials—while maintaining deep involvement in quantum and AI research communities through co-authoring technical surveys and engaging with venues like NeurIPS. Before Radical, he played a central role in scaling the Creative Destruction Lab where he championed building what is described as the world's first quantum software incubator and directly coached hundreds of founders, giving him unusually structured exposure to early AI teams compared to traditional VCs.

• A deep interest in applied artificial intelligence across diverse physical and biological sectors. Focuses on how machine learning solves real world problems, highlighting specific applications like World Labs bringing spatial intelligence to the physical 3D world, PrecisionAi enabling responsible herbicide use in agriculture, and software making protein structure prediction accessible in minutes.
• A strong focus on foundational AI infrastructure and core technologies. Emphasizes that compute and data are critical components alongside talent for advancing the space, regularly shouting out core AI companies like Cohere for natural language understanding and Deepliteai for powering efficient neural networks on edge devices.
• Enthusiasm for the evolution of search engines and language models. Frequently praises startups that challenge traditional search paradigms, pointing to YouSearchEngine as a prime example for breaking user records, opening platform access to developers, and shipping second product versions while competitors are still in the prototyping phase.
• Dedicated to the North American machine learning ecosystem with a distinct pride in Canadian innovation. Openly invites machine learning builders in Canada and the United States to connect, specifically celebrating Canada for stepping up its computing edge and cheering on Toronto based teams like the Acceleration Consortium for advancing material science.
• Seeking specific teleoperation infrastructure solutions. Put out a direct request for startups building tools for teleoperation, specifically looking for systems that can handle low complexity tasks with high latency right now and eventually scale to high complexity tasks with low latency.
• A high value placed on community building and academic networking within the artificial intelligence space. Regularly attends major industry conferences to meet founders and researchers in person, frequently mentioning plans to connect with old friends and make new ones at NeurIPS over consecutive years.