Invest at First Signal, Before Consensus Forms | Jamal Trotman, Radix Innovation Capital

What we talked about.

Jamal Trotman of Radix Innovation Capital joins host Sunny Ray to explain how his early stage venture fund invests at first signal, before consensus forms. Radix backs pre-seed and seed companies, often before product or revenue, by looking for signals that historically preceded venture-backable scale. Trotman describes his upbringing as an immigrant born in Guyana in a real estate development family, an early look inside Blackstone, and his mentorship with co-GP Sean Robinson that grew into a firm. He walks through a typical Tuesday morning of roughly 72 inbound decks, the importance of founder domain expertise and distribution, and Rachel, the firm's internal research operating system that preserves institutional learning. He also covers how Radix chooses LPs, its four to six week decision process, advice for first-time founders, sectors he finds underappreciated such as health tech and prop tech, why human judgment must stay in venture, and why patience and treating founders well matter over a decade-long horizon.

Radix's Jamal Trotman explains how to spot venture-backable founders before product or revenue, and why judgment must stay human.

The questions, and the answers.

What is Radix Innovation Capital and what problem are you solving?

Radix is an early stage venture fund investing at pre-seed and seed. Our whole thesis is investing at first signal. Many companies are pre-product and pre-revenue, so we look at the signals that have historically come before venture-backable scale, and we try to catch them as early as possible.

What does a first signal look like on a Tuesday morning?

On a typical Tuesday we might have about 72 inbound deals, and each deck takes three to five minutes to review. So we built internal infrastructure to spot signals. I look for deep founder domain expertise, a clear view of distribution and first customers, and what data moat or product learning the company can build.

Why did you build Rachel instead of just trusting your gut?

Rachel is our internal operating system for research. Venture suffers from entropy: when a GP leaves, the knowledge leaves too. Most GPs can't tell you why they passed on company 65. We can look across passes and investments and explain why, which also gives institutional LPs repeatability and auditability.

How do you choose which investors to take money from?

We don't raise from just anyone. We look at what our portfolio companies need, such as help with sales, tech, product, or go-to-market, and align LPs with those skills. Often they are high-network individuals with deep domain expertise who want to advise, so they are engaged rather than sitting on the sidelines.

Who has to say yes on a deal, and how fast do you move?

It takes a majority vote, so Sean and I both have to agree. About 9% of inbound deals make it to a first call, and under 4% of those become investable. Our process runs about four to six weeks, so we move pretty fast and keep founders informed.

What would you tell a first-time founder about getting in front of an early investor?

Be coachable. Every investor has a different lens, and one investor passing doesn't mean they're wrong or you are. Also build a community of founders who have built or are building, because they can make introductions to investors, as long as they believe you're building a venture-backable company.

Which sector are you excited about that consensus hasn't caught up to?

Not AI directly, since we are about to see its second order effects, much like smartphones led to Uber and Airbnb. I'm excited about health tech, including peptides and new regulation around replacing lab animal testing, and about prop tech, since housing demand is constant and AI should start shifting that space.

What part of venture should stay human as AI grows?

Judgment. You should never leave a robot or language model to decide where your judgment lies. Repeatable processes are great, but machines can't read a face, hear confidence in a voice, or capture emotion the way a person can. The final call has to stay with humans.