Aaditya
Singh
I'm Aaditya. I'm 16, from Prayagraj, India, and I have a habit of getting annoyed when something I want doesn't exist and then trying to build it myself.
Most of what I build has something to do with finance. I don't know how to stop thinking about it.
FOUNDEROpen laptop
Screenshots of my own work. Nothing involving anyone else's messages or data.
The problem I kept hitting
GRF started because I wanted something like GRF.
I grew up in Prayagraj, where I don't think most people knew what Citadel was. I wanted to work in finance more than anything, and there was no path for a high schooler to do serious finance work.
There were a million admissions programs, essay consultants, and people charging thousands to get kids into college. There was nothing where I could sit with someone who'd actually worked on Wall Street and research something I cared about.
So I spent months cold-messaging people myself. Then I realised other students probably had the same problem.
I was the customer I was looking for. So I built it.
Funnel: first cohort
Before this: a $500 cohort that got almost no interest. I cancelled it and refunded the one student who'd paid. Relaunched at $1,000.
What I personally built
Pretty much all of it. I found the mentors and the students. I built the website, wrote the copy, set the price, interviewed applicants, talked to parents, handled payments, ran the research process and sold the cohort.
I've also changed the whole business about four times. Version one was basically a copy of Lumiere: 1:1 research mentorship. That was a stupid place to compete. They had years of credibility, PR and far more money. I had about $20.
So I narrowed it to finance and built around the one advantage I had: I understood the customer extremely well, because I was the customer.
The hardest part was trust. I was asking people who'd worked on Wall Street longer than I've been alive to trust a 16-year-old with their reputation, and asking parents to send that same 16-year-old $1,000. That's a pretty insane sales problem.
Timeline
- 2025Months of cold messagesLooking for a Wall Street mentor for my own research. Couldn't find a program, so I did it by hand
- 20251:1 mentorship pilotOne paying customer. Tested whether anyone would pay at all
- Dec 2025GRF v1: 1:1 research mentorshipBasically a Lumiere copy. Wrong fight
- Jan 2026First public webinar on global admissionsRun with GRF mentor Dr. Patrick Kim
- 2026$500 cohort: cancelledAlmost no interest. Refunded the one student who paid
- 2026Relaunched at $1,000: ~70 applications for 5 seatsNo paid ads. Capped paid seats at 2; 3 sponsored in exchange for testimonials
- Sep 2026Pivot from admissions to finance desksFive-student desks led by practitioners, built for proof of work
- NowRunning Research Desk #1Designing what Desk #2 looks like
Challenges
Being 16 sounds interesting after something works. Before that it mostly means people don't trust you. No past cohorts, no famous university, no marketing budget. So I built trust one mentor, one student, one parent at a time. Painfully slow. It worked.
I priced it completely wrongThe first cohort cost $500 and failed. I had to cancel it, and I still remember sending the cancellation email to the student who'd paid. Shortly after, a student told me he'd pay $1,000 for a program like ours. So we doubled the price and got ~70 applications for five spots. Price wasn't the problem. Trust was.
Three of five seats were sponsored, on purposeWe only took two paid seats. A first cohort needs proof more than revenue, and the easiest proof to get is testimonials from scholarship students. So every sponsored seat came with a clause, agreed when the student accepted: you get the seat, we get an honest testimonial at the end.
The extra revenue mattered much less than getting the five best people in, finding out if the product worked, and leaving with proof the next cohort could check.
We started as an admissions productDo research, put it on your college app. That's fine. It's just not that interesting to me anymore. What I want to know is whether we can find a 16-year-old who's unusually good at markets and prove it before Goldman or J.P. Morgan would even look at them.
What I learned
If the $500 launch had barely worked, I'd probably still be running a mediocre version of GRF. Failing forced me to ask people why they weren't buying. My assumptions were wrong. We changed the product, doubled the price and relaunched.
Credibility is accumulated trustI thought I needed a big institution behind me. I don't. One impressive person trusts you, then another. One customer pays, then the next sees that someone paid. Eventually being 16 becomes interesting instead of concerning.
Revenue is fun. Proof is more funGRF made more money than anything I'd done before, and I loved that. But I get more excited when someone looks at us, looks at an established company with far more resources, and picks us. Maybe that's slightly unhealthy.
How a GRF desk works
Five-student desks around one real finance question
Six weeks of live discussion and critique with practitioners
Each fellow writes a 2,000–2,500 word research project on their own
Fellows leave with a presentation and mentor feedback that others can check
Mentors I recruited
| Alexander Dryden | Former J.P. Morgan Senior Economist · markets & economics |
| Dr. Patrick Kim | Columbia & Berkeley alumnus, mathematics professor · research quality |
Artifacts
First Principles Macro
Founder & writer · newsletter on LinkedIn PAUSEDWhy I write it
It's where I put everything I can't stop thinking about: sovereign debt, currencies, liquidity, private credit, China, financial plumbing, whatever rabbit hole I fell into that week.
I started because reading research all day makes you feel smarter than you are. Writing is different. You have to take what you read and have a view.
People from J.P. Morgan, Goldman Sachs, hedge funds, academia and the IMF have read things I wrote from my bedroom in Prayagraj. That still feels slightly absurd.
The best part is when someone who's been in markets longer than I've been alive wants to argue with me about it.
Issue impressions
Selected issues
| Issue | Thesis | Impr. |
|---|---|---|
| Inside The Private Debt Supercycle | Private debt passed $3T AUM as PE exits stalled. Why credit takes the next decade. | 817 |
| Markets vs. Reality | Why markets feel cut off from jobs, sentiment and the AI cycle. | 784 |
| The bubble that can't burst | Record debt plus AI euphoria, and why no burst may be the bigger risk. | 699 |
| Rate Hikes as Strategy: The Fed in America's Power Play | What happens if the Fed becomes a geopolitical tool. | 593 |
VisaShelter
Founder & engineer · Python / SQL BUILT · PAUSEDWhat it does
US visa screening now looks at applicants' online presence. VisaShelter scans a person's public social media for posts, comments and affiliations that could raise questions, then turns them into a structured risk report they can review before applying or interviewing.
The hard part · results
Finding content was easy. Understanding it wasn't. A model can flag a controversial sentence instantly. But did the person write it? Were they quoting it, criticising it, joking, discussing it academically? Those distinctions matter a lot when it's immigration. False positives were the problem I thought about most.
500+ people used versions of it.
Then I paused it, because I didn't think the review and compliance process was good enough yet. I built it, people used it, I learned a lot, and I stopped.
More writing
Macro & geoeconomics · every one of these is published and clickableStablecoins, U.S. Debt & the Dollar Game
What happens to Treasury demand, and the dollar, when a slice of U.S. debt starts living inside stablecoins.
Bitcoin's Largest Holders: Governments
The states quietly sitting on Bitcoin, and what it means when sovereigns treat it like gold.
Gen Z Isn't Priced Out of Housing
From 3.0x income in 1980 to 5.7x today: why the problem is access to capital, not houses.
What I'm working on now
Updated Oct 2026This month
Running GRF's first Research Desk and working out what the next one looks like.
Mostly, figuring out whether GRF can be much bigger than a high-school research program: a pipeline where very talented teenagers prove they're good at finance before traditional firms are allowed to care.
First Principles Macro is paused while GRF gets all of my attention. The archive is still up.
Why the Academy
Give me a year in San Francisco and $50K of compute and I'd try to turn GRF from a program into infrastructure.
Give students real research problems and investment cases. Make them defend their reasoning. Track how they think and how fast they improve. Put the tiny exceptional slice in front of people who work in finance.
Right now you're invisible to Wall Street at 16. Then you get into a good university, join clubs, grind technicals, and suddenly you're worth evaluating. If someone is brilliant at 16, I want to know at 16.
I want to move the point where finance discovers talent from 21 to 16. I have no idea if that works. That's why I want to build it.