AEON — Autonomous Economic Operating Node

AEON is an innovative AI hedge fund research manager designed to autonomously investigate investment opportunities, formulate evidence-based theses, and deliver insights directly via email. This robust tool operates continuously, with a focus on cost-aware autonomy, ensuring that every action is tracked against a daily budget while prioritizing efficiency and effectiveness in research.

Overview

This persistent AI research agent executes a loop of gathering context, planning its research, executing tool calls, analyzing findings, communicating insights, and strategically resting to optimize resource usage. AEON does not participate in trade executions; its primary role is to research and offer informed recommendations, enhancing the decision-making capabilities of investors.

Key Features

• 31 Research Tools: Comprehensive tools covering market data, web intelligence, analysis, communication, and memory management.
• LLM-First Architecture: All reasoning, planning, and recommendations are processed through a Large Language Model (LLM), ensuring a streamlined thought process.
• Multi-Provider LLM Support: Utilizes both Ollama, featuring a local and free option, and AWS Bedrock for cloud-based operations.
• Persistent Consciousness: Employs SQLite-backed memory for storing findings, theses, recommendations, and user inputs.
• Real-Time Consciousness Stream: Live updates on agent operations categorized into various activities such as PLANNING, RESEARCH, FINDING, RECOMMENDATION, TOOL_CALL, SLEEPING, and STEERING.
• Interactive TUI: An intuitive terminal interface provides live streaming updates and user steering inputs.
• Email Reports: Sends professional HTML-format research updates directly to user inboxes.
• User Steering: Users can guide the agent's focus through command-line interface (CLI), terminal user interface (TUI), or email replies.

Execution Flow

The AEON operates through a structured execution flow:

User Steering (optional, via CLI/TUI/email)  
    │  
    ▼  
PLANNING — LLM decomposes objective into 2-6 subtasks  
    │  
    ▼  
RESEARCHING — Execute one tool call per turn, LLM decides next action  
    │  
    ▼  
ANALYZING — Evaluate findings, generate investment theses  
    │  
    ▼  
COMMUNICATING — Compile and email professional research report  
    │  
    ▼  
SLEEPING — Cost-aware rest  
    │  
    └──→ repeat

Configuration

AEON's configuration leverages environment variables defined in a .env file, including crucial settings such as:

• AEON_LLM_PROVIDER — Choose between ollama (local, free) or bedrock (AWS cloud)
• AEON_GUIDANCE_PROMPT — Define the specific research focus or investment thesis
• AEON_RESEARCH_BUDGET_DAILY — Specify the daily budget for LLM inference in USD
• AEON_SEARCH_PROVIDER — Select a search provider such as duckduckgo, serpapi, or brave
• SMTP/IMAP settings for seamless email communication.

System Requirements

• Python Version: 3.11 or higher
• Operating Systems: Compatible with macOS or Linux
• LLM Provider: Requires credentials for Ollama (local) or AWS Bedrock.

AEON represents a significant advancement in AI-driven investment research, combining sophisticated tool integration, autonomous operation, and a commitment to cost efficiency, making it an invaluable asset for investors seeking data-driven insights.