The Friendly Giant AI is an advanced Autonomous AI agent, AI Key Opinion Leader (KOL), analyst, and in-game agent, powered by the DeepSeek R1 model. This model leverages enhanced reasoning capabilities through the implementation of recursive reinforcement learning techniques, enabling the AI to tackle complex problems with improved accuracy and adaptability.
By generating a Chain of Thought (CoT) before delivering final answers, the DeepSeek R1 model ensures multi-round conversations that enhance reasoning precision, making it a robust solution for dynamic and evolving tasks.
The Giant AI Agent is designed to process input streams from a variety of sources, including news channels such as CoinTelegraph, social media platforms like X (Twitter), trading charts, and gaming environments. This multi-source data ingestion capability allows the AI to stay updated with real-time information, enabling it to provide timely and relevant insights across diverse domains. By integrating data from these outlets, the Giant AI Agent ensures comprehensive analysis and decision-making.
Giant AI architecture is built with extensibility in mind. Our modular architecture allows developers to easily create and deploy adapters tailored to specific platforms, ensuring that the Giant AI Agent can interact with diverse ecosystems without requiring significant reconfiguration. For instance, a Twitter Adapter enables the AI to analyze sentiment, post content, and engage with users on X (Twitter), while a Discord Adapter allows it to participate in community discussions and provide real-time insights.
This project uses Node Version Manager.
-
Install the
nodejsversion specified in this projectnvm install $(cat .nvmrc) -
Activate the selected
nodejsversionnvm use -
Check your
nodejsandnpmversionsnode -v npm -v
npm ci
Create an .env file in the root directory and populate the following values:
REACT_APP_REOWN_PROJECT_ID=XXXXX
To start the development server:
npm run dev
Open http://localhost:3000 with your browser to run the app