This tool automates the production of financial research reports from inception to final draft. By providing a stock ticker, the system initiates an automated process that gathers financial statements, ownership structures, and relevant industry news. An AI engine then scrutinizes the raw data to compare peers, identify emerging trends, and highlight potential risks. The final output is a structured report featuring integrated charts, data visualizations, and a definitive investment recommendation. Reports are available in both Markdown and Word formats. Furthermore, a built-in workflow engine automates broader industry and macroeconomic research. The data analysis agent is capable of processing Chinese-language queries, generating its own analysis scripts, and producing graphs with proper Chinese labeling. It is designed specifically to make equity research both faster and more precise.
Multi-Source Data Aggregation: Automatically retrieves financial reports, shareholder structures, industry benchmarks, and company fundamentals.
Intelligent Financial Analysis: The AI performs comprehensive quantitative assessments, including growth rates, profitability margins, debt-to-equity ratios, peer benchmarking, and forward-looking estimates.
Automated Visualization: Dynamically generates professional-grade financial charts and data visualizations.
In-Depth Report Composition: Delivers a full-length equity research report complete with a specific investment thesis.
Modular Workflow Engine: Executes automated sequences for specialized industry studies and macroeconomic environmental scans.
The platform utilizes a five-layer architecture to transform raw data into professional reports.
Data Input Layer
Core Processing Layer
Workflow Engine
Report Generation Layer
Output Formats
1. Multi-Source Data Integration
akshare interface.2. Intelligent Analysis Engine
3. Professional Report Output
Requirements
Steps
Clone the repository:
git clone https://github.com/li-xiu-qi/financial_research_report
cd financial_research_report
Install the necessary dependencies:
pip install -r requirements.txt
Configure the environment:
Create a .env file in the root directory:
OPENAI_API_KEY=your_openai_api_key
OPENAI_BASE_URL=https://api.openai.com/v1
OPENAI_MODEL=gpt-4
Basic Report Generation
from research_report_generator import generate_report
target_company = "SenseTime"
target_company_code = "00020"
target_company_market = "HK"
generate_report(target_company, target_company_code, target_company_market)
Integrated Report Generation
from integrated_research_report_generator import IntegratedResearchReportGenerator
generator = IntegratedResearchReportGenerator()
generator.run_full_pipeline()
Industry Research Workflow
from industry_workflow import IndustryResearchFlow
from pocketflow import Flow
flow = Flow()
flow.add_node("industry_research", IndustryResearchFlow())
shared_data = {
"industry": "Artificial Intelligence",
"context": []
}
flow.run(shared_data)
akshare financial data ecosystem.Data Analysis Agent
Workflow Engine (PocketFlow)
Data Collection Tools
Quantitative Financial Analysis
Valuation Models
Risk Management
Generated reports typically include the following sections:
LLM Configuration The system is compatible with:
Analysis Parameters
DeepSeek-OCR WebUI: Batch OCR with Markdown Tables and Visual Bounding Boxes
Open English Dictionary: 25,000+ LLM-Refined Word Entries for Deeper Chinese Understanding
Tongyi DeepResearch: 30B Agent Model Beats GPT and Claude on Search Benchmarks
HackGPT Enterprise Review: AI-Native Pentesting for Security Teams
Alger Music Player: Play Grayed-Out NetEase Songs with Desktop Lyrics
Turso Database: A Rust-Based SQLite-Compatible Engine
ERPNext Open Source ERP: Installation Guide for Accounting and Inventory
SelfyAI: Build Your Own AI Agent as a Virtual World Asset
Build Web Apps Using Only SQL: A Guide to SQLPage
AI Peer Review Tool for Neuroscience: LLM-Driven Meta-Review Framework
Microsoft’s NLWeb: Converting Any Website into a Conversational Interface
ZeroSearch: Training LLMs to Search Without Real-World Search Engines