What is Perplexideez
Perplexideez is a local AI assistant that supports users in quickly searching for information on the web and in self-hosted applications. The Perplexideez project is based on a Postgres database, supports Ollama or OpenAI-compatible endpoints, and uses a SearXNG instance for web search. Perplexideez has multi-user support, single sign-on (SSO) functionality, and provides a beautiful embedded display of search results. Perplexideez aims to provide better self-hosted service integration and user experience.
Main features of Perplexideez
- Web search: Uses AI technology to help users filter web search results and improve search efficiency.
- Conclusion source tracking: Users can view the source annotations used by the LLM (large language model) and directly access the source code.
- In-depth learning: The LLM generates follow-up questions for users, making it convenient for users to further explore content they are interested in.
- Search tracking: Users can save search results they are interested in for easy later retrieval.
- Experience customization: Supports users in selecting appropriate models according to different tasks, and provides environment variables and UI configuration to optimize resource usage.
- Search sharing: Users can easily share search result links with others and perform access control.
Technical principles of Perplexideez
- Natural language query: Uses a large language model (LLM) to process and understand natural language queries, filter search results, and generate relevant follow-up questions.
- Supports integration with Ollama or OpenAI-compatible endpoints, which provide interfaces for AI models to perform complex language processing tasks.
- Database support: The project backend uses a Postgres database to store user data, search history, and preference settings, ensuring data persistence and multi-user support.
- Self-hosted search instance: Based on a SearXNG instance, it searches the web; SearXNG is an open-source, privacy-protecting metasearch engine.
- Multi-user and single sign-on (SSO): Supports multi-user use, with each user having an independent data view. Integrates OIDC (OpenID Connect) SSO, allowing users to access multiple services with a single identity authentication.
- Containerized deployment: Uses container technologies such as Docker and Kubernetes to enable flexible deployment of Perplexideez in various environments, including bare metal, virtual machines, and cloud platforms. Containerization also improves the portability and scalability of the application.
Project address of Perplexideez
- GitHub repository：https://github.com/brunostjohn/perplexideez
Application scenarios of Perplexideez
- Personal knowledge management: Individual users search and organize their personally saved information, quickly finding the materials they need.
- Enterprise internal search: Enterprises use it as an internal search engine to help employees quickly access enterprise knowledge bases, documents, and data.
- Research and academia: Researchers and scholars search academic resources, track research citations, and generate research questions.
- Education: Educational institutions assist teaching, helping students and teachers quickly obtain educational resources and academic materials.
- Development and programming: Developers search programming-related documentation, code snippets, and development resources to improve development efficiency.