← All projects
AIOpen SourcePythonAutomation

Open source

Polymarket Trading Bot

A prediction-market signal pipeline for Polymarket sports markets. It ingests markets, normalizes pricing inputs, scores edge and liquidity, and keeps execution gated until the model earns trust through review.

76

Unit tests passing across contract, ingestion, model, and risk

0

External dependencies in core model engine (pure stdlib)

Gated

Phase 3 execution disabled by default; validate first

Context

Turning market noise into signal

Prediction-market trading demands more than a clever signal. Sportsbook odds, CLOB depth, timing, liquidity, and risk state all have to resolve into an auditable decision before any execution step deserves authority.

Approach

An auditable ingestion-to-intent pipeline

Structured the pipeline around ingestion, normalization, model scoring, and disabled execution. The decision engine emits sized order intents with signal attribution, while the live-trading path stays off until paper review proves the model is worth trusting.

Outcome

From market discovery to sized intent

  • → A full path from market discovery to sized order intent.
  • → Auditable decision packets with signal attribution.
  • → A validation-first design that keeps live execution disabled by default.