Ellen Chisa brings an unusually technical perspective to venture investing, having co-founded Dark where she created an entirely new backend programming language with its own editor and infrastructure, giving her deep empathy for developers and practical insight into how new programming tools actually get adopted. As an "inception" investor at Boldstart, she focuses on meeting potential founders even before they decide to start companies and emphasizes whether they can realistically displace the largest incumbent over 5-10 years while running tightly scoped learning experiments rather than just writing code.

• Deep interest in the intersection of artificial intelligence and user experience, specifically regarding how humans interact with agents versus how agents interact with software. This includes a curiosity about future interface paradigms, such as how notification systems must evolve by 2026 to handle new workflows rather than just repeating current patterns.
• Focus on the practicalities and limitations of generative AI in daily workflows. There is an appreciation for founders who understand that managing AI currently feels like directing a very junior employee, as illustrated by a personal experiment using tools to create organization labels for a toddler's toy bin.
• Commitment to inception-stage investing, often looking to back founders at the very beginning of their journey before a product is fully formed. The criteria for this typically involves solving a clear, immediate pain point and having a team ready to tackle massive problems, with a willingness to write checks as early as two weeks after a major life event.
• Enthusiasm for deep tech and physical world applications, such as robotics and biotech. This includes excitement about advancements in dexterous manipulation, highlighted by a robot successfully handling Lego bricks, as well as models that can capture the movement of proteins rather than just their static states.
• Pragmatic view on business viability and market selection, particularly regarding the economics of AI. A key concern is whether teams are designing products that users love but are ultimately too expensive to run profitably, alongside the advice to ensure the target market is actually worth the effort of winning.
• Passion for developer tools and distinct go-to-market strategies. This involves an appreciation for starting with extremely narrow ideal customer profiles, such as targeting specific types of developers using open source packages, before expanding to a broader audience.