Polymarket, run like a portfolio.

Every Polymarket trader is scored on verified on-chain fills: total PnL, dollar-weighted win rate, Sharpe, Sortino, profit factor, max drawdown, and Kelly fraction. Those metrics roll up into a composite 0-100 Trader Score across five dimensions: Performance, Risk Management, Consistency, Edge, and Experience.

Every metric is sliced by category (Politics, Crypto, Sports, and more) and recency, from 24h to lifetime. A minimum of 20 closed positions is required before any score counts, so noisy track records don’t rank.

A filter defines which trades you care about: thresholds on any trader metric, market criteria like liquidity and days remaining, and trade criteria like size and entry price. Whitelists pin traders or markets in; blacklists keep them out.

When a filter matches, a pod acts: an email alert or a paper trade with full PnL simulation. A live estimate shows matched traders, markets, and trades per day before you commit.

Every paper pod carries its own risk envelope: starting capital, position size as a percentage of capital, max exposure per position, position limits per event and per trader, and hourly trade-rate caps.

Exits are configurable: follow sells, proportional exits, sell-at price targets, or hold to resolution. Every simulated position closes by a rule you chose. The discipline most traders skip, enforced by the system.

On-demand AI research on any market, event, or trader. Quick mode answers in seconds; Deep mode runs multi-source research with stronger reasoning; Dispute mode analyzes contested resolutions against UMA precedents. Every output is a confidence-scored assessment with key factors and sources.

Underneath the Dispute analysis, sits the resolution intelligence layer nobody else tracks: the full UMA lifecycle (Proposed, Disputed, Settled), hourly bulletin-board sync from on-chain adapter contracts, and line-by-line rule-change diffs on every market.