Frontier Research Fund

October 2026

Correspondence: hi@frontierresearch.fund

Abstract

Frontier Research Fund invests the first money in companies that begin as research. We back scientists and engineers at the frontier of their fields, and the people building the instruments, data, models, and automation those fields now share. We invest at the earliest stage, when the result is still the whole company.

Keywords: artificial intelligence, biology, chemistry, materials, climate, neuroscience, physics, research stack

Thesis

For most of a century the sciences moved at different speeds and spoke different languages. A result in one field took a decade to become a tool in another, if it ever did. It now takes about a year. A model trained on protein structures is a working instrument in materials science within a year. An automation platform built for chemistry is running biology the next. The datasets, the file formats, the compute, and increasingly the people are the same.

So the frontier is no longer seven frontiers. It is one, approached from seven directions, and the people who can see across it are starting companies. They look like labs for a while. Their first customers are often other researchers. Their advantage is a result, and the first investor has to be able to read it.

What we back

Artificial intelligence: especially as an instrument for the other sciences.
Biology: programmable cells, new measurement, and automation that takes the bench out of the loop.
Chemistry: computational discovery, self-driving laboratories, and synthesis at speeds a human bench cannot match.
Materials: new matter for energy, electronics, and structures, found by search rather than by chance.
Climate: capture, storage, and the grid, where a result in one of the sciences above changes what a kilowatt-hour or a tonne costs.
Neuroscience: interfaces, recording, and the models that make the brain's data legible.
Physics: fusion, quantum systems, sensing, and the instruments that open a new regime of measurement.
The research stack: what the seven fields now share, built by companies that serve all of them at once.

artificial intelligence biology chemistry materials climate neuroscience physics

instruments, data, models, automation, compute

Figure 1. Seven fields, one stack. A company that improves the stack moves every field above it; a company in any one field now begins by standing on it.

How we work

We write first checks. We read the paper before the deck. A preprint, a thesis chapter, or a prototype that works in one corner of a lab is enough to write to us with.

We need to understand what you found, why it is true, and why it is yours to build on. Once we invest we stay close: to the science, to the first hires, and to the question of when a result becomes a company.

Correspondence

If you are a researcher thinking about a company, or already running one, write to us. Tell us what you found, what it makes possible, and what you need next. Send the paper. We answer.

Limited partners and people who run labs reach us at the same address.¹