Orchard Experience Fuels the Launch of LunarBoost
Orchard Experience Fuels the Launch of LunarBoost
A conversation with Orchard alums Dylan Minor and Miles Suarez-Davis
A conversation with Orchard alums Dylan Minor and Miles Suarez-Davis
LunarBoost started the way a lot of great Orchard stories do: with curiosity, collaboration, and belief in an idea. Alums Dylan Minor ’22 and Miles Suarez-Davis ’22 sketched an equation that grew into a data company. Today they’re preparing an October 1 launch of a system that blends real-time market math with a custom LLM designed to capture hard-to-measure signals like founder quality and leadership durability. In back tests, they report beating the market by 200% to 1,600% in a given year. This year, they’ve also been approached by institutional investors, including firms with $10B+ in assets under management, who want to review their datasets.
What began with two Orchard students now includes nearly a dozen teammates studying at places like Princeton and Stanford. Their aim is to democratize high-quality investing information for both firms and everyday investors. Below, Dylan (senior at Park Tudor) and Miles (senior at University High School) explain how it works and why Orchard’s progressive education made them feel ready to build it.
Q: Start with the simple version, then take us deeper. What is LunarBoost?
Miles: We sell data, not advice. Our system turns thousands of signals into a clear score for each stock so people can make informed decisions.
Dylan: Under the hood, we built our own LLM with tight parameters to compile qualitative factors that typical screeners miss, like founder track record, ownership and incentives, and execution history. We calculate that right alongside real-time quantitative data like profitability, growth, and return on invested capital so the score reflects both the story and the numbers.
Miles: We sell data, not advice. Our system turns thousands of signals into a clear score for each stock so people can make informed decisions.
Dylan: Under the hood, we built our own LLM with tight parameters to compile qualitative factors that typical screeners miss, like founder track record, ownership and incentives, and execution history. We calculate that right alongside real-time quantitative data like profitability, growth, and return on invested capital so the score reflects both the story and the numbers.
Q: How did this go from an idea to a team and a product?
Miles: It literally started on a fishing trip. We were talking about AI, I sketched the math, and it turned into a very big equation and a lot of late nights.
Dylan: We began just the two of us and now the team includes people studying at places like Princeton and Stanford. Everyone brings different expertise (econ and quant, engineering, product, marketing), and that mix is baked into the system. It is not something you can Google. It is thousands of hours from people who like doing this in their free time.
Miles: It literally started on a fishing trip. We were talking about AI, I sketched the math, and it turned into a very big equation and a lot of late nights.
Dylan: We began just the two of us and now the team includes people studying at places like Princeton and Stanford. Everyone brings different expertise (econ and quant, engineering, product, marketing), and that mix is baked into the system. It is not something you can Google. It is thousands of hours from people who like doing this in their free time.
Q: What makes your approach different, and how is it performing so far?
Miles: If a quick search could tell you what wins, everyone would be rich. Our edge is the combination of quantitative finance with a structured qualitative layer from our LLM.
Dylan: In back tests, we have beaten the market anywhere from 200 percent to 1600 percent in a given year. Back tests are not guarantees, but the signal is strong enough that we keep widening the data and improving the model. We have also been approached by institutional investors, including firms with ten-billion-plus in assets under management, who want to review our datasets.
Miles: If a quick search could tell you what wins, everyone would be rich. Our edge is the combination of quantitative finance with a structured qualitative layer from our LLM.
Dylan: In back tests, we have beaten the market anywhere from 200 percent to 1600 percent in a given year. Back tests are not guarantees, but the signal is strong enough that we keep widening the data and improving the model. We have also been approached by institutional investors, including firms with ten-billion-plus in assets under management, who want to review our datasets.
Q: Where does Orchard show up in all of this?
Miles: Orchard didn’t teach me to fit in a box. It taught me to innovate and take the next step beyond. That mindset is our day to day.
Dylan: I learned to try and fail, then try again with full effort. That is what startups are. At Orchard, we were encouraged to lead, to experiment, and to take smart risks. That makes it feel normal to chase a hard problem before the world says you are old enough.
Miles: Orchard gave us opportunities to do things in middle school and high school that many people do not see until college. That confidence to build with other people is the whole reason LunarBoost exists.
Miles: Orchard didn’t teach me to fit in a box. It taught me to innovate and take the next step beyond. That mindset is our day to day.
Dylan: I learned to try and fail, then try again with full effort. That is what startups are. At Orchard, we were encouraged to lead, to experiment, and to take smart risks. That makes it feel normal to chase a hard problem before the world says you are old enough.
Miles: Orchard gave us opportunities to do things in middle school and high school that many people do not see until college. That confidence to build with other people is the whole reason LunarBoost exists.
Q: Who is this for, and what is the plan?
Miles: Two audiences. For B2B partners like hedge funds, private equity, and venture, we provide bulk datasets to power their strategies. For retail investors there is a monthly software service with a set number of lookups that shows scores and the factors behind them. The goal is to democratize high-quality investing information by making it more readily available.
Dylan: We can do this in the stock market because it is regulated and rich with history, which lets us back test and improve responsibly. We launch October 1 and will keep iterating after release.
Miles: Two audiences. For B2B partners like hedge funds, private equity, and venture, we provide bulk datasets to power their strategies. For retail investors there is a monthly software service with a set number of lookups that shows scores and the factors behind them. The goal is to democratize high-quality investing information by making it more readily available.
Dylan: We can do this in the stock market because it is regulated and rich with history, which lets us back test and improve responsibly. We launch October 1 and will keep iterating after release.
Note: LunarBoost provides data, not investment advice. All investing involves risk, and back-tested performance does not guarantee future results.
Comments