Finance & Trading

PolyHuntr

Prediction Market Analysis Platform

A real-time prediction market monitoring system that cross-references market contracts with news feeds to identify potential mispricings through proprietary probability assessment.

PolyHuntr hero image

// the problem

Challenge

Prediction markets react to news within minutes, making manual monitoring ineffective. Building an edge detection system requires real-time market data ingestion, multi-source news aggregation, natural language understanding for sentiment analysis, and probability modeling to identify significant deviations.

// what we built

Solution

PolyHuntr maintains real-time connections to prediction markets for live price feeds while aggregating news from dozens of sources. A proprietary ML pipeline analyzes news sentiment and assesses probability implications. When market prices deviate significantly from assessed probabilities, the system triggers multi-channel alerts.

// shipped

Key features

  • Real-time contract price streaming
  • Multi-source news aggregation
  • Proprietary probability assessment
  • Configurable alert thresholds
  • Position tracking dashboard
  • Historical performance analytics
  • Custom watchlist management
  • Telegram/Slack/Discord bots

// stack.json

Tech stack

The exact tools shipping this product in production.

  • Python
  • TypeScript
  • AI/ML
  • Market APIs
  • PostgreSQL
  • Redis
  • WebSocket
  • Next.js

// system.diagram()

Architecture

Real-time prediction market analysis with proprietary edge detection and multi-channel alerts

Prices Articles Markets Edge Log Next.js Dashboard frontend Market Scanner backend News Aggregator service Edge Detector ai Alert Engine service PostgreSQL database Redis cache
  • frontend
  • backend
  • service
  • ai
  • database
  • cache

// receipts

Results

  • Real-time prediction market integration
  • Multi-source news aggregation
  • AI-powered sentiment and probability analysis
  • Near-instant news-to-alert pipeline
  • Multi-channel alert delivery
  • Historical backtesting framework
Real-time market streaming
Data Feed
Dozens of aggregated feeds
News Sources
Proprietary probability model
Analysis
Near-instant news-to-alert
Latency

// next()

Have a project like this?

We build production systems with the same engineering rigor you see here. Let's talk.