PhantomAI exists because we lost money before we made money. The system you use today is the one we wished existed while we were blowing up our own accounts.
We started trading prediction markets like everyone else: following Discord tips, chasing favorites, averaging down on losers, turning a good day into a bad month. The account graph was a slow bleed that occasionally spiked green.
Every loss got written down. The postmortems stacked up. Patterns emerged. A team priced at 94¢ is a trap, not a lock. A back-to-back road favorite isn't a sure thing. Tank teams upset. Expert consensus is sometimes just loud agreement rather than signal.
When the notes became a system, the system became software. PhantomAI is the compiled version of that journey — every rule traces back to a trade that once cost us real money.
We don't put faces on this page because the system is bigger than any one person. The edge comes from the method, the discipline, and the machine — not a headshot.
Built the rulebook from a four-figure trading loss. Oversees product direction, method integrity, and the boring decisions that compound into long-term outperformance.
The multi-LLM consensus layer: Gemini, Groq, Sonnet, Opus. Every published signal survives a unanimous vote. Disagreement among models is itself a signal.
Rotating advisors across prediction markets, sports analytics, and quantitative research. They review the rulebook quarterly and stress-test new patterns before they reach production.
We publish two to four signals a day, never fifty. When the edge is absent, we send nothing. Patience is the product.
We track and publish every result — wins and losses both. No cherry-picking, no deleted screenshots. The record is the record.
Consistent, rule-based execution at the right prices is what separates winners from gamblers. We operate as if the money is scarce — because it always is.