An AI-powered chatbot designed to assist developers building on the NEAR blockchain, built with Next.js, Vercel AI SDK, OpenAI, Exa Search, and NEARBlocks API.
Features · Tech Stack · Getting Started · Deployment · Roadmap · Chat History Analysis
- Dynamic access to up-to-date NEAR documentation and resources via Exa Search
- Integration with NEARBlocks API for real-time blockchain data
- Persistent chat history using Vercel KV
- Authentication with NextAuth.js
- Responsive design with Tailwind CSS
- Next.js with App Router
- Vercel AI SDK for streaming chat UI
- OpenAI API (GPT-4o model)
- Exa Search for real-time access to NEAR documentation
- NEARBlocks API for blockchain data
- Vercel KV for data storage
- NextAuth.js for authentication
- Tailwind CSS for styling
- shadcn/ui for UI components
- Node.js 18+
- pnpm
- OpenAI API key
- Exa Search API key
- NEARBlocks API key
- Vercel account (for deployment and Vercel KV)
-
Clone the repository:
git clone https://github.com/your-repo/near-founder-copilot.git cd near-founder-copilot
-
Install dependencies:
pnpm install
-
Set up environment variables: Copy
.env.example
to.env.local
and fill in the required values:cp .env.example .env.local
-
Run the development server:
pnpm dev
-
Open http://localhost:3000 in your browser.
Deploy your own instance of NEAR Founder Copilot to Vercel:
-
Build the application:
pnpm build
-
Start the production server:
pnpm start
- Integrate NEAR-specific documentation and resources via Exa Search
- Implement latest GPT-4o model for improved performance
- Integrate NEARBlocks API for real-time blockchain data
- Enhance context management for more coherent multi-turn conversations
- Implement fine-tuned AI model specifically for NEAR development
- Add support for code completion and debugging assistance
- Implement multi-language support
- Create a plugin system for extending functionality
- Develop a feedback mechanism for continuous improvement of responses
- Implement advanced caching strategies for faster response times
The NEAR Founder Copilot includes a chat history analysis feature that provides insights into user interactions and common queries. This analysis helps improve the copilot's responses and identify areas for further development.
- Data Collection: Chat histories are stored securely using Vercel KV.
- Analysis Scripts: Python scripts (
analyze_chats.py
andreformat_text.py
) process and analyze the chat data. - Visualization: The analysis generates visualizations of top user queries and interaction patterns.
- Identification of common user questions and topics
- Trends in NEAR development interests
- Areas where additional documentation or resources may be needed
We take user privacy seriously. All chat data is anonymized before analysis, ensuring that individual users cannot be identified. The data is kept strictly confidential and is used solely for improving the NEAR Founder Copilot's performance and understanding broader trends in NEAR development queries. We do not share or expose any user-specific information, maintaining the trust and privacy of our developer community.
The chat history analysis contributes to the continuous improvement of the NEAR Founder Copilot, ensuring it remains a valuable tool for the NEAR developer community while respecting user privacy.
We welcome contributions to the NEAR Founder Copilot! Please see our Contributing Guide for more details.
This project is licensed under the MIT License - see the LICENSE file for details.
- Vercel for the Next.js framework and Vercel AI SDK
- OpenAI for the GPT-4o model
- Exa for providing real-time access to NEAR documentation
- NEARBlocks for blockchain data API
- NEAR Protocol for the blockchain platform
Built with ❤️ by the NEAR community