virattt/dexter

An autonomous agent for deep financial research

What it solves

Dexter is an autonomous financial research agent designed to handle complex financial queries. It eliminates the manual effort of gathering and analyzing market data by automating the research process, from planning to final answer synthesis.

How it works

Dexter uses a loop of task planning, execution, and self-reflection. It decomposes a user's complex question into a structured research plan, selects and executes the appropriate tools to gather real-time financial data (such as income statements, balance sheets, and cash flow statements), and then validates its own work to refine the results until a confident, data-backed answer is reached.

Who it’s for

It is built for individuals interested in financial research and analysis, though the project explicitly states it is for educational and informational purposes and not for professional investment advice.

Highlights

  • Autonomous Planning: Automatically breaks down complex queries into step-by-step research tasks.
  • Real-Time Data: Integrates with financial datasets and web search (via Exa or Tavily) to access current market information.
  • Self-Validation: Iterates on its research steps and checks its own work to ensure accuracy.
  • WhatsApp Integration: Allows users to interact with the agent via a WhatsApp gateway.
  • Evaluation Suite: Includes a built-in system using LangSmith and an LLM-as-judge approach to track and score the agent's correctness.

Related

  • Project
    AI4Finance-Foundation/FinRobot

    An open-source AI agent platform for financial applications that automates investment research and algorithmic trading by combining LLMs with deterministic quantitative analytics.

  • Project
    u14app/deep-research

    An AI-powered research tool that automates the gathering of information from the web and local files to generate comprehensive, in-depth reports.

  • Project
    OpenBMB/XAgent

    An autonomous LLM-driven agent that solves complex tasks by breaking them down into subtasks and executing them safely within a Docker-sandboxed environment.

  • Project
    assafelovic/gpt-researcher

    An open-source deep research agent that automates web and local document analysis to produce detailed, factual research reports with citations.

  • Project
    Autumn-27/ARTEX

    An AI-driven autonomous penetration testing system that uses a multi-agent architecture and dual-graph tracking to automate vulnerability discovery and complex attack chains.