Setup and Usage
After launching the application, please configure all large language model-related parameters properly to ensure the system can connect to the large model service you have chosen.
In addition to LLM, you also need:
- Tavily: API key, used for Tavily web search, application address: https://www.tavily.com/
- Bocha: API key, used for Bocha multimodal search, application address: https://open.bochaai.com/
To experience the full functionality, none of the above parameters can be missing.
Introduction
"BettaFish" is an innovative multi-agent public opinion analysis system built from scratch. It helps break information cocoons, restore the original public sentiment, predict future trends, and assist decision-making. Users only need to raise analysis needs like chatting; the agents automatically analyze 30+ mainstream social platforms at home and abroad and millions of public comments.
Features
- AI-Driven Omnipresent Monitoring: An AI crawler cluster operates 7x24 hours, comprehensively covering over 10 key domestic and international social media platforms such as Weibo, Xiaohongshu, TikTok, and Kuaishou. Not only does it capture hot content in real-time, but it also drills down into Massive user comments, allowing you to hear the most authentic and widespread voices of the public.
- Comprehensive Analysis Engine Beyond LLM: Equipped with 5 specialized Agents, combined with middleware such as fine-tuned models and statistical models, it ensures the depth, accuracy, and multi-dimensional perspective of analysis results through collaborative multi-model operation.
- Powerful Multimodal Capabilities: Breaking through text and image limitations, it can deeply analyze short video content from platforms like Douyin and Kuaishou, and accurately extract structured multimodal information cards such as weather, calendars, and stock information from modern search engines, enabling you to fully grasp public opinion dynamics.
- Agent "Forum" Collaboration Mechanism: Endowing different Agents with unique toolsets and thinking patterns, introducing a debate moderator model, and facilitating chain thinking collisions and debates through the "forum" mechanism. This not only avoids the cognitive limitations and homogenization caused by single-model thinking and communication but also fosters higher-quality collective intelligence and decision support.
- Seamless Integration of Public and Private Domain Data: Not only analyzing public opinion but also providing high-security interfaces to support the seamless integration of your internal business databases with public opinion data. Breaking down data barriers to provide powerful analytical capabilities for vertical businesses with "external trends + internal insights."
- Lightweight and Highly Scalable Framework: Based on a pure Python modular design, achieving lightweight and one-click deployment. The code structure is clear, allowing developers to easily integrate custom models and business logic for rapid platform extension and deep customization.