AI founders: here’s what we want to know
Simon Hopkins, who leads AA’s screening process for us (AKA our “dealflow guy”), writes…
We’ve recently made several changes to our online application form, the first stage for founders looking to pitch to Angel Academe investors. Most of the changes were iterative responses to questions and issues that have come up in our weekly screening meeting. But one area sees an entirely new set of questions: AI.
It’s hardly a surprise that a substantial number of companies now coming our way are either AI native businesses or otherwise feature AI as a substantial part of their value proposition (just shy of a quarter of the startups that have applied to us in 2026, in fact). To help us understand exactly how these companies are deploying AI, and the implications of doing so, our application form now features these additional questions for relevant companies:
AI Model: Which model(s) do you use, or have you built your own?
AI training and verification: Briefly explain how you train your model and, where appropriate, how you verify results.
AI compute costs: How do you meet the costs involved in training and running your model?
Other AI issues and considerations: Are there any other implications of your approach to AI (eg. data collection, confidence level requirement, privacy, scaling)?
At this stage, we aren’t looking for detailed responses (less than100 words in fact), and we aren’t trying to "catch out" applicants! Rather, we're just trying to get an early sense of how founders are approaching AI, and in the process preparing them for some of the more detailed questions that may arise on a subsequent screening call.
So... if you’re the founder of an early stage company doing exciting work with AI (at whatever “depth”), please apply to us here. And if you have questions about the process, feel free to drop me a line at simon@angelacademe.com!
(My thanks to the screeners who worked on the revised form, especially John Souter (who led the process), Angela Cretu and Audrey Mandela.)