CrowdWisdomTrading Prediction Market AI Agent

Project Overview

This project implements a multi-agent AI system designed to research events and generate predictions related to prediction markets.

The system collects real-world information from the web, stores contextual knowledge, simulates trader analysis, and generates a final prediction.

The architecture is designed to mimic how traders analyze news events before making market decisions.

---

Architecture

The system is composed of multiple agents working together:

1. Scraper Agent

   • Collects event information from the web using Apify Google Search Scraper.
   • Extracts titles and descriptions of relevant events.

2. RAG Memory Layer

   • Stores collected event information.
   • Retrieves knowledge for downstream analysis.

3. Trader Agents

   • Simulate market traders analyzing event information.
   • Generate sentiment signals (positive / negative / neutral).

4. Decision Agent

   • Aggregates trader signals.
   • Produces a final prediction about market direction.

---

System Workflow

Web Events ↓ Scraper Agent (Apify) ↓ RAG Knowledge Memory ↓ Trader Sentiment Analysis ↓ Decision Agent ↓ Market Prediction

---

Tech Stack

• Python
• Apify (Web scraping)
• Modular multi-agent architecture
• Simple RAG memory system

---

Installation

Clone the repository:

git clone https://github.com/yourusername/prediction-ai-agent.git

Navigate into the project folder:

cd prediction-ai-agent

Create virtual environment:

python -m venv venv

Activate environment:

Mac / Linux: source venv/bin/activate

Install dependencies:

pip install -r requirements.txt

---

Configuration

Create a file named config.py and add your Apify API token:

APIFY_TOKEN = "your_apify_token"

EVENT_QUERY = "AI startup funding news"

---

Run the Project

Run the AI agent:

python main.py

---

Example Output

Starting AI Prediction Agent...

Starting Prediction Agent Researching event: AI startup funding news

Storing knowledge Retrieving knowledge Traders analyzing knowledge

FINAL RESULT Prediction: Market uncertain

---

Possible Improvements

Future improvements may include:

• Integration with real prediction market APIs (Polymarket / Kalshi)
• LLM-based reasoning using OpenRouter
• Improved RAG with vector databases
• Real trader wallet analysis
• Market category classification (sports / politics / weather)

---

Author

Darshan S BE Computer Science & Data Science PES College of Engineering