Appearance
Octagon Agents for Market Intelligence
Octagon provides a variety of specialized AI agents, each designed for specific investment and market research domains. This page provides an overview of each available agent.
Agent Categories
Prediction Markets Agents
Our prediction markets agent provides research on Kalshi events by surfacing price drivers, comparing market vs. model probabilities, and identifying potential mispricings.
| Agent | Description |
|---|---|
| octagon-prediction-markets-agent | Researches prediction market events with cache and refresh modes for speed or freshness |
Public Market Agents
Our public market intelligence agents provide comprehensive analysis of publicly traded companies through SEC filings, earnings call transcripts, stock data, and financial metrics.
| Orchestrator | Specialized Agent | Description |
|---|---|---|
| octagon-agent | octagon-sec-agent | Analyzes SEC filings data |
| octagon-transcripts-agent | Analyzes earnings call transcripts | |
| octagon-stock-data-agent | Analyzes stock market data | |
| octagon-financials-agent | Analyzes financial statements | |
| octagon-holdings-agent | Analyzes institutional ownership and holdings data | |
| octagon-crypto-agent | Analyzes cryptocurrency market data and digital assets |
Deep Research Agent
Our deep research agent provides powerful capabilities for extracting data from websites and conducting comprehensive market research.
| Agent | Description |
|---|---|
| octagon-deep-research-agent | Conducts in-depth research |
Best Practices for Agent Queries
- Be Specific: The more specific your query, the more accurate the response will be. Include time periods, specific metrics, or companies when relevant.
- One Task Per Query: Focus each query on a single analytical task for best results.
- Provide Context: When needed, include relevant context to help the agent understand your requirements.
- Use the Right Agent: Select the most appropriate agent for your specific task rather than using a general-purpose agent.
Using Agents in Multi-Step Workflows
For complex research tasks, you may want to combine multiple agents. For example:
- Use the octagon-sec-agent to analyze a company's latest filings
- Use the octagon-financials-agent to review financial statement trends
- Use the octagon-transcripts-agent to compare management commentary with reported results
For advanced multi-agent workflows, consider using the OpenAI Agents SDK which provides tools for agent orchestration.