https://github.com/ashutoshrajgupta/recipe-generator-chatbot-with-rag
This project demonstrates how to build a recipe generator chatbot using Retrieval-Augmented Generation (RAG). The chatbot can fetch recipes from a local database or perform web searches to provide users with accurate and relevant cooking instructions.
https://github.com/ashutoshrajgupta/recipe-generator-chatbot-with-rag
astradb groq langflow openai python3 retrieval-augmented-generation
Last synced: 7 months ago
JSON representation
This project demonstrates how to build a recipe generator chatbot using Retrieval-Augmented Generation (RAG). The chatbot can fetch recipes from a local database or perform web searches to provide users with accurate and relevant cooking instructions.
- Host: GitHub
- URL: https://github.com/ashutoshrajgupta/recipe-generator-chatbot-with-rag
- Owner: AshutoshRajGupta
- Created: 2025-01-05T15:11:14.000Z (9 months ago)
- Default Branch: main
- Last Pushed: 2025-01-19T14:36:45.000Z (9 months ago)
- Last Synced: 2025-01-19T15:31:45.920Z (9 months ago)
- Topics: astradb, groq, langflow, openai, python3, retrieval-augmented-generation
- Language: Python
- Homepage: https://astra.datastax.com/langflow/62f7f0d1-631b-449d-a892-b475e8114ed1/flow/a17833e2-2c48-4065-800e-839a451c1d2b
- Size: 19.3 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Recipe Generator Chatbot with RAG (Retrieval-Augmented Generation) using Langflow
**"AI Recipe Master: A Chatbot with RAG-Powered Recipe Retrieval"**
## Overview
This project demonstrates how to build a recipe generator chatbot using Retrieval-Augmented Generation (RAG). The chatbot can fetch recipes from a local database or perform web searches to provide users with accurate and relevant cooking instructions. This step-by-step guide uses Langflow to simplify the integration of multiple components like chat inputs, agents, and databases.
---
### **What is RAG (Retrieval-Augmented Generation)?**
Retrieval-Augmented Generation (RAG) is an advanced AI approach that combines the strengths of information retrieval systems and generative AI models. It works by retrieving relevant documents or data from a knowledge base and using them to generate precise and contextually accurate responses.
#### **How RAG Works**
1. **Input Query** : The user provides a query (e.g., "How do I make garlic bread?").
2. **Retrieval Step** : The system searches a knowledge base (e.g., a database or document store) for relevant information.
3. **Augmentation Step** : The retrieved information is provided as context to the generative AI model (like GPT) to generate an enhanced, accurate response.
4. **Response Generation** : The AI generates an answer based on the retrieved context and the user query.
#### **Advantages of RAG**
* **Accuracy** : It provides factual responses grounded in the retrieved data.
* **Efficiency** : Reduces hallucinations by anchoring responses in actual information.
* **Scalability** : Can integrate with large-scale databases, such as AstraDB, for rapid information retrieval.
---
### **AstraDB: A Cloud-Native NoSQL Database**
**AstraDB** is a fully managed database-as-a-service built on Apache Cassandra, designed for high performance, scalability, and flexibility. It’s an excellent choice for implementing the retrieval step in a RAG system.
#### **Key Features of AstraDB**
1. **Scalability** : Handles large volumes of data and high query loads efficiently.
2. **Cloud-Native** : Runs seamlessly on AWS, Google Cloud, or Microsoft Azure.
3. **APIs for Flexibility** : Supports GraphQL, REST, and gRPC APIs for easy integration with various applications.
4. **Serverless Architecture** : Scales automatically based on demand, reducing operational overhead.
5. **Free Tier** : Offers a generous free tier, making it accessible for development and testing.
---
### **Why Use AstraDB in a RAG System?**
1. **Data Storage** :
* Recipes, ingredients, steps, and metadata can be stored in structured tables.
* AstraDB's columnar data model supports efficient querying and retrieval of recipe information.
1. **Fast Retrieval** :
* AstraDB provides low-latency data access, crucial for real-time AI applications like chatbots.
1. **Integration** :
* Its REST and GraphQL APIs simplify connecting to AI workflows created in Langflow or custom Python scripts.
1. **Scalability for Large Datasets** :
* AstraDB can handle extensive recipe datasets, including millions of rows, without performance degradation.
1. **Data Synchronization** :
* Real-time updates allow the database to stay up-to-date, ensuring the chatbot fetches the latest recipes.
## Features
- **User-Friendly Interaction** : Chat-based interface for recipe queries.
- **RAG Integration** : Combines database retrieval and generative AI for robust recipe fetching.
- **Dynamic Search** : Automatically switches to web search if the recipe is not available in the local database.
- **Database Integration** : Uses AstraDB to store and retrieve recipe data efficiently.
- **Customizable Agents** : Configurable agents to handle information retrieval and processing.
---
## Prerequisites
Before starting, ensure you have the following installed:
- **Python 3.8+**
- **Langflow** : [Langflow Installation Guide](https://github.com/logspace-ai/langflow)
- **AstraDB Account** : Sign up at [AstraDB](https://www.datastax.com/astra)
- **API Access** : Gemini API or any web-search API for external information retrieval.
- **Other Libraries** : Install dependencies listed in `requirements.txt`.
---
## Setup Instructions
### 1. Clone the Repository
```bash
$ git clone https://github.com/your-username/recipe-generator-rag.git
$ cd recipe-generator-rag
```
### 2. Install Dependencies
```bash
$ pip install -r requirements.txt
```
### 3. Configure Environment Variables
Create a `.env` file in the root directory and add the following:
```env
ASTRADB_ID=
ASTRADB_SECRET=
WEB_API_KEY=
LANGFLOW_API_KEY=
```
### 4. Setup AstraDB
- Create a database named `recipes`.
- Add a table with fields for recipe name, ingredients, and steps.
---
## Step-by-Step Workflow
### 1. Setup
- Sign up or log in to Langflow.
- Create a new project flow and name it `Recipe Generator`.
### 2. Creating Inputs/Outputs
- **Inputs** : Add a chat input node to capture user queries.
- **Outputs** : Add a text output node to display chatbot responses.
### 3. Agent Configuration
- Use Langflow to create agents.
- Connect tools for database retrieval and web search API integration.
### 4. Database Integration
- Link the chatbot to AstraDB using Langflow's database tool.
- Ensure queries to AstraDB return accurate recipe details.
### 5. Functional Testing
- Query the chatbot with example inputs like:
- "How do I make a chocolate cake?"
- "What's a good recipe for pasta salad?"
- Validate responses from both the database and web search.
---
## Usage
### Run the Chatbot
```bash
$ streamlit run ./main.py
```
### Example Query
**User** : "How do I make garlic bread?"
**Chatbot** : "To make garlic bread, you will need: ... [recipe details fetched from database or web]"
---
## Screenshots
### Langflow Architecture
Here is the visual representation of the Langflow workflow used to build the chatbot:
![Langflow Architecture]
### Chatbot Output
Example of chatbot interaction in the Streamlit UI:
---
## Langflow Playground
Explore the Langflow playground with a live demo:
[Langflow Playground Demo](https://astra.datastax.com/langflow/62f7f0d1-631b-449d-a892-b475e8114ed1/flow/a17833e2-2c48-4065-800e-839a451c1d2b)
---
## Acknowledgements
- **Langflow** for simplifying AI workflow management.
- **AstraDB** for robust database services.
- **OpenAI** for powering generative responses.
---
## Contributing
Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.