Matt Murphy is a venture capital investor at Menlo Ventures who previously spent over 15 years at Kleiner Perkins, where he co-led the pioneering iFund focused on iPhone apps and backed companies like DocuSign and AppDynamics. He's known for his specific benchmark that $2M seed rounds should generate roughly $1M in ARR to indicate early product-market fit, and his focus on vertical SaaS for tech-laggard industries like construction and insurance, along with AI-first infrastructure that enhances developer productivity.

• Championing Anthropic's dominance in the enterprise sector, frequently citing data to argue that their models have surpassed competitors like OpenAI in corporate adoption. He highlights their massive revenue growth and significant market share gains as proof that they are the preferred choice for business applications.
• Focus on AI-powered developer tools and the democratization of coding, looking for platforms that allow non-engineers to build software or that streamline workflows for existing developers. He expresses enthusiasm for startups like Lovable and Cursor that combine code generation with review processes to transform how software is built.
• Excitement for foundational models applied to robotics and the physical world, moving beyond just digital chatbots. He specifically points to SkildAI as an example of a company successfully building general-purpose brains for robots, viewing this as the next major frontier in artificial intelligence.
• Reliance on proprietary market research and data-driven insights to validate investment theses. He frequently references Menlo Ventures' internal reports on generative AI, healthcare trends, and consumer behavior to demonstrate deep understanding of market shifts and adoption rates.
• Appreciation for strong, high-trust co-founder dynamics, particularly noting the effectiveness of complementary skills and low ego in leadership teams. He points to the sibling duo behind Anthropic as a prime example of the type of aligned partnership he values.
• Emphasis on the importance of clean data and context for unlocking model capabilities. He argues that simply having powerful models is insufficient without the right workflows and high-quality data, showing interest in startups that provide the infrastructure to help labs improve model intelligence.