A venture capital firm investing at the inception stage, in founders solving high-stakes problems in data-rich industries.
Prompt Capital backs founders whose undeniable obsession with a problem compels them to go solve it. They are building in markets where the software must work, where getting it wrong carries real consequences - sometimes regulated, always unforgiving. As these companies scale, the record and data their products accumulate in production becomes the moat. If this sounds like you, we'd love to hear from you.
From Insights
All insights
Every venture investment has two parts: an implied complaint about the current state of the world and a deeply optimistic belief that the complaint can be remedied.
Read articleBuilding an agent that can call a few tools is no longer a hard problem. The hard problem is keeping that system accurate and dependable once it's handling real customer data, in production, inside an industry that doesn't tolerate mistakes. We invest in founders who deeply understand this, and will stop at nothing to succeed.
Model quality stops being the differentiator quickly. What doesn't transfer is the working record a company builds once real customers run real work through it: every edge case seen, every correction made, every audit passed. In regulated markets that record is also the sales pitch - the next customer is buying proof the system has already survived their world. A better base model helps the incumbent with a record more than it helps a challenger without one.
In operationally complex industries, general-purpose models attempt work they can't be trusted to finish. That gap doesn't close from the generalist side, because closing it means absorbing the rules, the audit trail, and the liability - exactly what horizontal tools are built to avoid. The companies that close it from the inside own the category.
We see the same pattern in every sector we study: the moment the work carries real obligations - an audit to pass, a certification to hold, rules that change at every state line - horizontal tools stop being enough. What looks like red tape from the outside is the barrier to entry.
We write the first check at inception. We like to meet founders before there's a deck. We don't just expect something unpolished, we prefer it. We invest in businesses solving high-frequency, high-stakes problems, that humans hate to do. If you're building something you've been obsessing over for years, we want to hear from you. Email us here.
Managing Partner
Loren leads Prompt Capital. She was previously a partner at Bowery Capital, where she now serves as a venture partner. She is based in San Francisco and focuses on AI-native companies in regulated, data-rich industries.
Co-Founder and Venture Partner
David Schellhase is a co-founder and venture partner at Prompt Capital. He was the first general counsel at Salesforce, Groupon, and Slack. He is lead independent director at Okta and co-teaches a class in the MS&E department at Stanford.
Companies turning industry knowledge into systems of action. Companies appear here once they are ready to be public.
An AI-powered pre-construction operating system helping commercial general contractors turn fragmented project information into structured scoping, bid evaluation, and pricing workflows.
An AI-native insurance distribution company addressing structural inefficiencies in the Excess & Surplus commercial insurance market.
A field service management platform for the agentic era, beginning with the complex and under-automated pest-control industry.
An intelligence platform for public-company CFOs and investor relations teams, built around proprietary communications, market, media, and research data.
Are you working on a problem that you're completely obsessed with solving? We'd love to hear about it. We read everything that comes in.