melih-unsal/DemoGPT

🤖 Create LLM agents in a second with your prompts. Everything you need to create an LLM Agent - tools, prompts, frameworks, and models - all in one place.

What it solves

DemoGPT streamlines the creation of LLM-based applications and AI agents. It removes the need to manually write complex LangChain pipelines and boilerplate code by automatically generating the necessary code to build interactive Streamlit applications from user instructions.

How it works

DemoGPT uses a multi-step process to transform user instructions into a functional app:

  1. Planning: It generates a high-level plan based on the user's request.
  2. Task Creation: The plan is broken down into specific, actionable tasks.
  3. Code Generation: These tasks are converted into individual code snippets.
  4. Assembly: The snippets are combined into final code that powers a Streamlit application.

Additionally, the DemoGPT AgentHub provides a library of pre-built tools (like Wikipedia, Python, and YOLO object detection) and a framework for creating custom agents (such as ToolCallingAgent and ReactAgent) that can combine these tools with RAG (Retrieval Augmented Generation) for complex reasoning.

Who it’s for

Developers and AI enthusiasts who want to quickly prototype and deploy AI agents and LLM applications without writing extensive manual code.

Highlights

  • Automatic App Generation: Converts user instructions into interactive Streamlit apps via LangChain.
  • AgentHub: A comprehensive suite of tools including search, weather, Python execution, and computer vision (YOLO).
  • Flexible RAG: Supports multiple vector stores including Chroma, Pinecone, and FAISS for document-based reasoning.
  • Reasoning Agents: Includes a ReactAgent that provides a transparent, step-by-step reasoning process for multi-step tasks.