How I Use AI to Supercharge My Due Diligence on Private Investments
Why artificial intelligence is becoming an essential assistant for early-stage deal research
When it comes to public markets, research is straightforward. There are investor presentations, quarterly filings, analyst reports, and earnings calls—an ocean of information to swim through.
But private investments are a different beast. Often, the only document available is a slick investor deck designed to raise capital—not to answer your toughest questions.
I don’t have a research team. But I do have something nearly as powerful: AI.
Over the past year, I’ve started to delegate parts of my due diligence to large language models. And today, I want to share where this works brilliantly—and where I still roll up my sleeves myself.
To make this practical, I’ll walk you through a real-world example: a company called Astroteq AI, which is trying to forecast major earthquakes by analyzing cosmic rays. As wild as that sounds—it might just work.
AI Is the Best Way to Learn a New Industry—Fast
Often, my first challenge is simply understanding what an early-stage company actually does. That’s especially true when it comes to deep tech—like quantum computing, or in the case of Astroteq, predictive earthquake technology.
So I start with the basics.
I ask AI to explain the industry in simple terms, covering:
What the company’s technology is based on
What the industry is trying to solve
The main trends, opportunities, and roadblocks
For this kind of foundational research, Google Gemini has become my go-to. (Yes, I use premium versions of both Gemini and ChatGPT for a fair comparison.)
In my experience, Gemini is like a junior analyst who really wants to impress—it gives me a broader, deeper overview than ChatGPT, which sometimes delivers only the key facts but misses the connective tissue that ties them together.
In the case of Astroteq, I learned that earthquake forecasting is generally split into three categories:
Early warning
Probabilistic forecast
Deterministic prediction
Astroteq aims to lead in the latter—deterministic forecasting—which is the holy grail of the field.
I then asked Gemini to dive deeper into that specific niche. It returned a report titled: "The Final Frontier: A Strategic Analysis of the Deterministic Earthquake Prediction Market." That kind of second-layer insight is incredibly helpful in framing my own investment view.
AI Makes Peer Group Research Far More Efficient
After understanding the business model and tech, I usually want to understand who else is in the space. This is often straightforward for public peers, but far harder when the landscape consists of private competitors.
This is another area where AI shines.
By asking both ChatGPT and Gemini to surface comparable companies, I often get a richer result set than I would manually. They each bring different names to the table—so using both gives me broader coverage.
That said, here’s the caveat: Once I have my peer group, I still do the comparison myself.
Neither model (yet) delivers meaningful, apples-to-apples competitor comparisons. Every industry has its quirks—and what matters in one sector might be irrelevant in another. Maybe I haven’t nailed the perfect prompt yet, but for now, I still rely on good old-fashioned reasoning to line up valuation, differentiation, and traction.
When I find a better method, you’ll be the first to know.
Bottom Line: AI Is My Research Wingman—But I’m Still in the Pilot Seat
AI is no replacement for conviction. But it is a brilliant tool to:
Kick off research on unfamiliar industries or tech
Understand the broader landscape in plain English
Identify potential peers (especially in private markets)
I’ve included both reports I commissioned for Astroteq so you can see what this process looks like in action.
Report 1 on the general industry (generated by Gemini): Download here
Report 2 on Deterministic Prediction (generated by Gemini): Download here
The verdict? AI gives me a faster, stronger start to any due diligence cycle. And that, in a world where speed and insight are alpha, is a real edge.
If you’ve cracked the code on using AI for peer analysis—or want to collaborate to find the best prompts—drop me a message. I’m confident we’ll solve that piece too, sometime in the next 12 months.
👍🏼👍🏼