Skip to content

Arm you with all the essential tools to integrate AI seamlessly into your apps, regardless of the coding language you're comfortable

Notifications You must be signed in to change notification settings

GiovanniSmokes/NexaAIOne

 
 

Repository files navigation

🧞‍♂️ NexaAIOne

Welcome to NexaAIOne, a centralized RESTful API hub for Artificial intelligence (AI). Designed for every developer. NexaAIOne platform brings advanced features and customizability right to your fingertips.

What is NexaAIOne?

In simple terms, NexaAIOne is a wrapper for OpenAI API that adds multiple essential capabilities, such as Memory, Caching, Document Q&A, and more.

Basic example for using NexaAIOne via API from your application (check documentation for more):

curl https://localhost/api/v1/app/1/1/chatgpt \
    -H "Content-Type: application/json" \
    -H "Authorization: Bearer $AUTH_TOKEN" \
    -d '{
    "cachingPeriod": 60, --> cache AI answer for 60 minutes
    "session": "user-1397", --> Define unique session ID for every user to have different memory & cache management 
    "fakeLLM": 0, --> If you would like to use fakeLLM (during development & testing), or you want this request to be routed to OpenAI
    "enableMemory": "shortMemory", --> Do you want to enable conversation tracking? Turning this on will retain a record of past conversations.
    "memoryOptimization": "summarization", --> Which memory management method you want to use (noOptimization, truncate, or summarization)
    "collection_id": 33, --> Use documents from collection id 33 to answer user question
    "userMessage": "How can I subscribe to your service?" --> send user question to NexaAIOne
}'

High Level Design of NexaAIOne

Features:

  • RESTful API All AI services are configured to be consumed as RESTful API, this way you can use them in any application you want.
  • Memory Management: Enhance your LLM requests with contextual memory, leveraging strategies from truncating old messages to embedding and summarizing conversations.
  • Collections (Retrieval-augmented generation (RAG)): create your own AI chat that answers from your own enterprise documents.
  • Caching Management: Improve response times and conserve tokens with efficient caching mechanisms.
  • Ready AI Services: Engage with AI for chats, audio, images, document chat.
  • Developing & Testing: Efficiently debug AI requests, use the "Fake LLM" AI interface, and ensure no wastage of AI tokens.
  • Troubleshooting & Debugging: Monitor and inspect all your API requests for a smoother troubleshooting experience.
  • Custom APIs: Design bespoke APIs tailored to each AI service.
  • Auto-API-Documentation: Seamlessly generates comprehensive documentation for all APIs, ensuring clarity and ease of use for developers at every skill level.
  • Transparent Costs: Crafted to minimize AI token expenses without hidden prompts or costs.
  • Track Usage: Gain insights into API requests, token usage per application, and more.

Supported AI Services

  • OpenAI ChatCompletion: Creates a model response for the given chat conversation.
  • OpenAI Transcription: Transcribes audio into the input language.
  • OpenAI Auto Translation: Translates audio into English.
  • OpenAI DALL·E: an AI system that can create realistic images and art from a description in natural language.
  • Microsoft Azure OpenAI
  • TranslateGPT
  • Text Classification
  • Summarize Text
  • Sentiment analysis
  • Support Agent able to search knowledge base and suggest opening ticket if now answer found
  • Chat with your Documents (Create Chatbot Agent for Sales,Support,HR...etc)

Documentation & Getting Started:

Installation

Getting Started

Feel free to contribute, suggest features, or join us on this journey to making AI accessible and efficient for all developers.

About

Arm you with all the essential tools to integrate AI seamlessly into your apps, regardless of the coding language you're comfortable

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • PHP 79.5%
  • Blade 15.5%
  • HTML 2.5%
  • Dockerfile 2.2%
  • Other 0.3%