4 years of resolved market data — included

Build the bot. Own the edge.

I built a Polymarket scanner that tracks large wallets and flags where positioning moves ahead of price. Now I'll show you how to build your own — and how to find an edge that's yours, not mine.

Access

Three ways in.

Start with the prompt stack and have something running tonight. Or take the whole build with the dataset behind it.

Founding cohort. 1-on-1 build calls are capped at 12 slots a month so they stay useful — when the month fills, new circle members start the following cycle. Current price is locked for as long as you stay subscribed.

Tier 01

The Build Stack

The prompt sequence that turns an empty folder into a working scanner.

$13.99
One-time · Instant access
  • The architecture prompt — hands Claude the full skeleton: ingest, wallet-level position tracking, trigger logic, alerting
  • The debug chain — follow-up prompts in the order things actually break: rate limits, malformed responses, timestamp drift, silent failures
  • Mutation prompts — swap the data source or change the trigger condition without rebuilding from scratch
  • The module map — what every piece does and why, so you can modify it instead of hoping it keeps working
Get the stack
Most popular Tier 02

The Blueprint

The full build, start to live — plus four years of resolved market data to test against.

$29.99
One-time · Includes Tier 01
  • Four years of resolved Polymarket history with wallet-level activity — the part you can't get anywhere else
  • The complete build on camera — nothing skipped, nothing edited out
  • Going live — API keys, wallet setup, hosting it so it runs with your laptop shut
  • Backtesting your logic — measuring calibration, checking edge decay year over year
  • Finding your own edge — the questions to interrogate a market with before you write a line of code
Get the blueprint
Everything Tier 03

The Inner Circle

Where builds get pressure-tested — and turned into something you can sell.

$69.99/month
Monthly · Cancel anytime
  • Everything in Tier 01 and 02, plus new builds and datasets as I ship them
  • 1-on-1 build calls — screen share, we fix your specific failure
  • The room — builders running live bots, comparing what's holding up and what's decaying
  • Package and sell your own build — infrastructure, pricing, delivery, support. Your product, your customers.
  • Live in 30 days — or we keep working with you until it is
Join the circle
The moat

Anyone can build a bot. Almost nobody can test one.

This is the difference between a scanner that looks clever and one you'd actually size into.

4 YRS

Resolved Polymarket history with wallet-level activity, cleaned and ready to run your logic against.

A sample big enough to trust

Most people build a bot, watch it two weeks, and can't separate skill from variance. Thousands of resolved markets across four years is enough to know whether a signal is real.

Every outcome already known

Markets settle to 0 or 1. No ambiguity, no survivorship games. You can measure calibration directly — something you can't do cleanly in equities or spot crypto.

Edge decay, made visible

Run your logic on 2022, then run it on 2025. If it dies somewhere in between, you found a crowded signal before it cost you a dollar.

Verify the thesis yourself

Don't take my word that certain wallets lead price. Check it against the history, measure the lag, and decide whether it's worth building on.

Build path

Four states, one bot.

These aren't chapters — they're the states your bot actually passes through. You'll know exactly which one you're in.

Skeleton

It runs

The prompt stack gives Claude the architecture. You get a scanner that pulls data and prints output. Rough, but yours.

Backtest

It's measured

Point your logic at four years of resolved markets. Find out if what you built has an edge or just a story.

Live

It's running

Keys, wallet, hosting. It watches markets whether or not you're at your desk, and tells you when something fires.

Owned

It's yours to sell

Inside the circle: how to package what you built, price access, and deliver it — as your product, to your customers.

The guarantee

Your bot goes live in 30 days — or we keep going until it does.

I'm not promising you profit; nobody honest can. What I'll promise is the thing I control: if you're inside the circle, follow the build, and your bot isn't running in 30 days, we stay on it together until it is.

Get started
Portrait
Who's teaching this

I built it because nobody would show me.

I'm not a fund and I'm not selling signals. I'm a builder who got tired of paying for black boxes that stopped working the month after I subscribed — so I learned to build my own, and I kept the receipts.

Everything here is the actual process: the prompts I use, the mistakes that cost me weeks, and the dataset I test against before I trust anything. If you can follow instructions and ask a model good questions, you can build this.

See the tiers
FAQ

Straight answers.

No — but you need to be willing to read what the model writes and understand roughly what it's doing. The prompts do the heavy lifting. The module map exists so you're not copy-pasting blind. People who've never shipped anything have gotten a scanner running; people who refuse to look at the code tend to get stuck at the first error.

Resolved Polymarket markets with their price history and wallet-level activity — cleaned into a format you can query and backtest against. The point isn't the file size; it's that every market in it has a known outcome, so you can score your logic honestly instead of guessing.

No. You get the architecture and the method, then you point it wherever you want. That's deliberate — if a thousand people run the identical trigger on the identical market, the edge is gone by month two. Most of the Blueprint is about finding a condition that's yours.

To build and backtest: nothing beyond the tier price and an API key. Trading capital is a separate decision entirely, and it's yours to make. Plenty of people run the scanner for weeks in observation mode before they put a dollar anywhere.

No, and you should be skeptical of anyone who does. Markets change, edges decay, and execution is on you. What I guarantee is delivery: the prompts, the build, the data, the calls — and inside the circle, a running bot in 30 days or we keep working until you have one.

Tier 01 gets you a scanner running tonight. Tier 02 adds the full build, the going-live setup, and the four years of data to test against — it's the one most people want. Tier 03 adds me: calls, the room, new drops, and how to turn your build into a product of your own.

Immediately after checkout. One-time tiers deliver instantly; the circle gives you room access and your first call slot right away.