At Precedent.vc, My Co-GP is an AI Copilot

I review a lot of early-stage companies. More than I can reasonably hold in my head, more than I can pattern match instinctively, and more than I can evaluate without bias creeping in, no matter how disciplined I try to be. That is not a unique problem. It is the reality of modern venture. Like many of us who have been in this game for a long time, we don’t have an access problem. We have a signal problem. Especially when you are a solo GP running a lean team.

Early-stage investing still runs on a familiar playbook. Pattern recognition, intuition, and fast judgment under uncertainty. That works until it does not. Those same instincts push us toward what we have seen before. They reward familiarity. They quietly filter out things that do not fit the mold. Even the best investors are not immune to this. I was not.

Because my first love was engineering, I built an AI co-pilot to sit alongside my process, not to replace judgment but to structure it. Every company I review is tracked and scored across a set of dimensions, such as founder-market fit, distribution advantage, workflow-replacement potential, market size, and category timing. Each input is weighted, and every company rolls up into a score from 0.0 to 5.0. The model evolves as I see more companies and more outcomes. It is not static. It compounds.

Most of my deals come in via email, so I simply forward those emails to my AI co-pilot.

What changed for me was not speed. It was clarity. Instead of reacting to individual companies, I started seeing patterns across categories, repeated signals across founders, and early indicators of what is actually working. Just as important, I started to see what I might have been missing.

When inputs are structured and evaluated consistently, you can remove a surprising amount of the noise that influences decisions. It is similar to a blind resume test. When you remove identifying signals and focus on the underlying attributes, different candidates rise to the top. The same thing happens here. Every company is evaluated against the same framework, the same criteria, the same weighting. Not who the founder is, not where they went to school, not how polished the pitch is. What matters is what actually drives outcomes. That shift changes what surfaces. Over time, it reveals patterns that would otherwise be easy to overlook.

The outcome is not more deal flow. It is better triage. Instead of asking if something is interesting, I am asking if it meets the bar relative to everything else I am seeing. That is a different filter, and it is a much higher one.

This has also changed how I communicate what I am seeing. I have started producing a detailed quarterly report for my LPs that aggregates deal flow, scoring patterns, and emerging market trends. It is one thing to see individual companies. It is another to step back and understand what is consistently rising to the top and why. That layer of insight has become just as valuable as the deals themselves.

I am starting to share a small slice of those insights more broadly through Precedent.vc. Not everything, but enough to give a sense of where conviction is forming and how the underlying patterns are shifting. For those who want a more direct view into the companies themselves, I have created a separate stream called Precedent Signal. It is intentionally small and focused, a monthly cut of what I am seeing where I have real conviction.

dealscoreai.xyz

I also decided to open up a piece of my AI Copilot, because when we share learnings, it makes the system better for everyone.

I have found my system works best when you can calibrate across a set of deals you are seeing, not just one at a time. The real value comes from comparing signals and understanding how something stacks up relative to everything else you are seeing.

The future of Investing

I do not think AI replaces investors. But I do think it changes how good investors operate. There is less guesswork, more structure, and better pattern recognition over time. The edge is no longer just access. It is how you process what you see.

I am curious how others are thinking about this…

Are you still relying on instinct, or are you starting to build systems around your decision-making?

What’s one signal you trust when evaluating early-stage companies that others might overlook?

Where do you think bias shows up most in early-stage investing?

PS: Interested in building in this space? Drop me a note. I have a few fun projects underway, and I’m looking to work with interns and to find a Principal who is excited about this direction for the venture.