Devecho is an AI-powered social media assistant designed to generate engaging LinkedIn posts based on user-provided content. It automates data collection, retrieval, and post-generation using AI-driven techniques and multiple autonomous AI agents that collaborate to optimize the content. This combination of intelligent agents and advanced retrieval methods enables users to create high-quality content efficiently.

📺 Watch Devecho in action:

• Automated Data Collection: Fetches relevant information from URLs, GitHub repositories, or topics using specialized AI agents.
• Retrieval-Augmented Generation (RAG): Leverages AI agents to analyze retrieved data and generate contextually accurate responses.
• Post Optimization: Uses AI to fine-tune content for engagement, clarity, and tone.
• LinkedIn Integration: Allows users to post directly from the assistant.
• Telegram Bot Support: Provides a conversational interface for users to interact with the system.

shaswatsingh3101-devecho/
├── deployment/
│   ├── knowledge_base.py        # Handles data collection from URLs, GitHub, and topics
│   ├── knowledge_retrieve.py    # Implements RAG pipeline and AI agent coordination for content retrieval
│   ├── linkedin.py              # LinkedIn API interaction (posting, user details)
│   ├── main.py                  # Orchestrates the end-to-end pipeline
│   ├── post_gen.py              # Generates social media posts with optimal engagement using AI agents
│   ├── telegram_bot.py          # Telegram bot interface for user interaction
│   ├── tone_config.py           # Manages different writing tones
│   ├── Procfile                 # Defines process to run the Telegram bot
│   ├── Readme.md                
│   └── requirements.txt         # List of dependencies
├── app/
│   ├── knowledge_base.py        # Handles data collection from URLs, GitHub, and topics
│   ├── knowledge_retrieve.py    # Implements RAG pipeline and AI agent coordination for content retrieval
│   ├── linkedin.py              # LinkedIn API interaction (posting, user details)
│   ├── main.py                  # Orchestrates the end-to-end pipeline
│   ├── post_gen.py              # Generates social media posts with optimal engagement using AI agents
│   ├── telegram_bot.py          # Telegram bot interface for user interaction
│   ├── tone_config.py           # Manages different writing tones
│   └── requirements.txt         # List of dependencies
├── assets/                      # Contains media assets for the project
└── Readme.md 

Check deployment dir for deploying your telegram bot

git clone https://github.com/SHASWATSINGH3101/devecho.git
cd devecho

pip install -r requirements.txt

• Add your Tavily, FireCrawl, Groq, and LinkedIn API keys to the environment:

export TAVILY_API_KEY="your_api_key"
export FIRECRAWL_API_KEY="your_api_key"
export GROQ_API_KEY="your_api_key"
export LINKEDIN_ACCESS_TOKEN="your_api_key"
export BOT_TOKEN="your_telegram_bot_token"

• Alternatively, modify the .env file and include the above keys.
• How to get the LINKEDIN_ACCESS_TOKEN:- link

Ensure you have the following dependencies installed:

pip install requirements.txt

python main.py

python telegram_bot.py

1. Start the bot with /start.
2. Use /new to begin post generation.
3. Follow the prompts to provide content.
4. Generate posts and upload them via /upload_linkedin.

To post content directly to LinkedIn:

python linkedin.py

Devecho supports multiple tones for post-generation:

• Professional: Clear and authoritative.
• Casual: Conversational and engaging.
• Educational: Explanatory and structured.
• Persuasive: Compelling with strong CTAs.

Change the tone using:

python tone_config.py professional

or through the Telegram bot with /tone.

• More advanced post structuring and style customization.
• Enhanced user feedback loop for refining generated content.
• Integration with additional social media platforms (e.g., Twitter, Facebook).

We welcome contributions! Fork the repo, create a new branch, and submit a pull request.

This project is licensed under the MIT License.

To cite this repository in publications:

@misc{CSVision,
  author = {SHASWATSINGH3101},
  title = {DevEcho},
  year = {2025},
  howpublished = {\url{https://github.com/SHASWATSINGH3101/DevEcho}},
  note = {GitHub repository},
}

For inquiries or support, connect via:

• 💬 Discord: shaswat_singh