Author: Mansoor Khan (GitHub: rebase-master)
Repository: agentic-ai-apps
This repository contains a collection of agentic AI applications and prototypes that explore how AI agents can collaborate, reason, and perform complex workflows using various frameworks and techniques — such as LangChain, LangGraph, and custom multi-agent orchestration.
Each branch in this repository represents a self-contained AI project, built to demonstrate different problem-solving capabilities — from order handling and customer service to summarization and product Q&A systems.
Each project lives on a separate branch to keep implementations modular and independent.
| Branch | Description |
|---|---|
customer-service-ai-agent |
A conversational customer service AI agent capable of handling user queries, managing support tickets, and simulating real-world customer interaction workflows. |
multi-agent-systems-with-langgraph |
A multi-agent orchestration system built using LangGraph, where different specialized agents (Product, Orders, Small Talk) collaborate to process user requests intelligently. |
orders-chatbot-with-custom-agent |
A custom AI agent chatbot that manages product orders — capable of updating, tracking, and summarizing order data through conversational inputs. |
product-qna-agentic-chatbot |
An Agentic Q&A chatbot designed to answer user questions about products using a retrieval-augmented approach (RAG) and structured reasoning. |
summary-chatbot-with-reviewer |
A summarization chatbot enhanced with a reviewer agent that validates and refines generated summaries for higher factual accuracy. |
- 🧩 Agentic architecture: Each project demonstrates autonomous or semi-autonomous AI behavior.
- ⚙️ Multi-agent coordination: LangGraph and tool-based routing to dynamically select specialized agents.
- 📦 Modular design: Independent branches for clarity and maintainability.
- 🗣️ Conversational workflows: Integration with chat-style user interfaces and context-aware prompts.
- 🔍 Retrieval and reasoning: Use of embeddings, context memory, and structured tool invocation.
- LangChain / LangGraph
- OpenAI APIs
- Python (async / FastAPI / tool decorators)
- Pandas, Pydantic, and data processing utilities
- JSON-based state and workflow orchestration
Each branch includes:
- A Jupyter Notebook (
.ipynb) or Python module demonstrating the agent logic. - Example prompts and workflows.
- README or inline documentation describing how the agent functions.
To explore a project:
# Clone the repository
git clone https://github.com/rebase-master/agentic-ai-apps.git
cd agentic-ai-apps
# Switch to a specific project branch
git checkout <branch-name>This repository is open source under the MIT License. Feel free to explore, adapt, and build upon these agentic AI systems.
Mansoor Khan 🔗 (GitHub Profile)
If you find these projects useful, consider giving the repository a ⭐ on GitHub!