Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/krik8235/rag-agent-system-chrome-extention
A Chrome extension that uses AI agents to help users negotiate the best deal
https://github.com/krik8235/rag-agent-system-chrome-extention
fastapi koyab llama3 pydantic pydantic-models python3 replit together-ai
Last synced: about 1 month ago
JSON representation
A Chrome extension that uses AI agents to help users negotiate the best deal
- Host: GitHub
- URL: https://github.com/krik8235/rag-agent-system-chrome-extention
- Owner: krik8235
- License: mit
- Created: 2024-11-21T14:39:34.000Z (about 1 month ago)
- Default Branch: main
- Last Pushed: 2024-11-21T17:16:12.000Z (about 1 month ago)
- Last Synced: 2024-11-21T17:38:34.958Z (about 1 month ago)
- Topics: fastapi, koyab, llama3, pydantic, pydantic-models, python3, replit, together-ai
- Language: Python
- Homepage:
- Size: 145 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Overview
A Chrome extension that uses AI agents to help users negotiate the best deals.
[Test website](https://rag-agent-system-chrome-extention.vercel.app/)
## Table of Contents
*generated with [DocToc](https://github.com/thlorenz/doctoc)*
- [Table of Contents](#table-of-contents)
- [Overview](#overview-1)
- [Key Features](#key-features)
- [Technologies Used](#technologies-used)
- [Project Structure](#project-structure)
- [Setup](#setup)
- [Usage](#usage)
- [Development](#development)
- [Package Management](#package-management)
- [Customizing AI Agents](#customizing-ai-agents)
- [Modifying Prompts](#modifying-prompts)
- [Contributing](#contributing)
- [Troubleshooting](#troubleshooting)
## Overview
A Chrome extension that uses AI agents to help the user negotiate the best deals:
1. **Context & Goal Setting**:
- From the chrome extention, ask users about their goals of the deal.
- Use Pydantic to define the schema via models
2. **Negotiation Starts**:
- Employ Llama 3.1 (running on Together AI) and call `/browse` API endpoint via FastAPI
- Record the log on the negotiation.log to track the conversation
3. **User Interaction**:
- Presents the seller's response and ask for a counter response.
- Allows the buyer to choose whether to accept or renegotiate.
4. **Response Generation**:
- Collect user input.
- Passes the information back to Llama 3.1 on Together AI.
- Generates response to the end user.
## Key Features
- Collect user inputs via Chrome extention
- Use Pydantic to build a schema and validate the inputs
- AI-driven context analysis & negotiation converation using Llama 3.1 on Together AI
- Call API via FastAPI and render response on the user-friendly interface
- Automated generation of response drafts using Llama 3.1
## Technologies Used
[data-doc-management]
- Upstage: Converts natural language queries into SQL [Doc](https://console.upstage.ai/docs/getting-started/overview)
- Chroma DB: Vector database for storing and querying standard contract clauses
- SQLite: Database for storing application data
[ai-model-curation]
- Together AI: Hosting Llama 3.1 for text processing, clause segmentation, and response generation
[deployment-framework]
- Python: Primary programming language
- Poetry: Backend build system and package installer/resolver [doc](https://python-poetry.org/docs/)
- Pydantic: Data validation library for Python
- Fast API: Web framework for building APIs based on standard Python type hints
- Replit: Deploy Fast API on Replit server (See the `.replit` and `replit.nix` for its configuration.)
- Koyeb: Deploy the backend repo and access in the production env
## Project Structure
```
.
└── frontend - react app
|
└── src
└── __init__.py
└── api.py = connect with database
└── main.py = build fastAPI connection
└── Prompts
└── Browse.py
└── Negotiate.py
└── Validate.py
└── tests
└── __init__.py
└── test_endpoint.py: test connection to the browser endpoint
└── test_negotiation.py: test and log the negotiation process and outcome
├── pyproject.toml
└── poetry.lock
< Koyeb deployment>
└── requirements.txt
└── Procfile
└── runtime.txt
```
## Setup
1. Install the `poetry` package manager:
```
curl -sSL https://install.python-poetry.org | python -
poetry --version
```
2. Install dependencies on pyproject.toml:
```
poetry install
```
3. Set up environment variables:
Create a `.env` file in the project root and add the following:
```
TOGETHER_API_KEY=your_together_api_key
DATABASE_URL=your_database_url
API_ROUTE=your_public_api_route
```
## Usage
1. Test the negotiation process and outcome on terminal:
```
poetry run neg
```
The outcome will be recorded in the negotiation.log file.
2. In a separate terminal, test the connection to the client interface:
```
poetry run client
```
The results will be printed on your terminal.
When you update the `main.py`, run this command to see if the update is working properly.
The backend will be available at `http://localhost:8000`.
3. Deploy the API via Koyab
```
koyeb deploy . /src --instance-type=free
```
*See the [doc](https://www.koyeb.com/docs/build-and-deploy/deploy-project-directory) for the Koyeb CLI installation and more.
Now you can access the API endpoint from the public URL `...koyab.app`.
4. Run the frontend. (Open another terminal and set the root to 'frontend')
```
npm i
npm start
```
The backend will be available at `http://localhost:3000`.
## Development
### Package Management with poetry
- Add a package: `poetry add `
- Remove a package: `poetry remove `
- Run a command in the virtual environment: add a script to pyproject.toml, then `poetry run `
- To reinstall all the package: after deleting poetry.lock file, run `poetry install`
[poetry cli](https://python-poetry.org/docs/cli/#add)
### Customizing AI Agents
To modify or add new AI agents, edit the `test_negotiation.py` file to test on the terminal, and/or edit the `main.py` file to apply the modification to the API.
### Modifying Prompts
Necessary prompts are stored in the `Prompts` folder. Add or modify the prompt accordingly.
(To avoid errors, we recommend adding a new prompt file to the Prompt folder.)
## Contributing
1. Fork the repository
2. Create your feature branch (`git checkout -b feature/your-amazing-feature`)
3. Commit your changes (`git commit -m 'Add your-amazing-feature'`)
4. Push to the branch (`git push origin feature/your-amazing-feature`)
5. Open a pull request
## Troubleshooting
Common issues and solutions:
- API key errors: Ensure all API keys in the `.env` file are correct and up to date.
- Database connection issues: Check if the Chroma DB is properly initialized and accessible.
- Memory errors: If processing large contracts, you may need to increase the available memory for the Python process.
- Issues related to the AI agents: Check the `negotiation.log` file for detailed error messages and stack traces.