⚡ BlogBolt AI : Topic In, Blog Out.

An AI-powered blog writing workspace that researches, plans, writes, and illustrates complete Markdown blogs, built with Streamlit and LangGraph.

· Features · How It Works · Quick Start · Troubleshooting

---

📖 Overview

Writing a good technical blog means researching, outlining, drafting, and creating visuals. BlogBolt AI automates that entire pipeline. Enter a topic, pick the audience and tone, and a multi-stage LangGraph workflow produces a publish-ready article with sources and AI-generated diagrams.

✨ Features

Generation

• 🧠 Structured blog planning and writing with Google Gemini
• 🔀 Smart routing: the app decides if a topic is closed-book, hybrid, or needs current information
• 🌐 Optional web research with Tavily, with sources included in the output
• ⚙️ Parallel section writing through LangGraph workers, merged by a reducer

Visuals

• 🎨 AI-generated technical images via Pollinations
• 🔤 Exact text labels on images using Pillow, so no garbled AI text

Workspace

• 📊 Live workflow progress and event logs
• 🗂️ Tabs for Article Preview, Markdown Source, Plan, Evidence, Images, and Logs
• 🕘 Recent blogs in the sidebar
• 📥 Export as .md or as a .zip bundle (blog + images)

🔄 How It Works

flowchart LR
    A[Topic + Audience + Tone] --> B{Router}
    B -->|Needs research| C[Research<br/>Tavily]
    B -->|Closed-book| D[Orchestrator<br/>Blog Plan]
    C --> D
    D --> E1[Worker 1]
    D --> E2[Worker 2]
    D --> E3[Worker N]
    E1 --> F[Reducer<br/>Merge Sections]
    E2 --> F
    E3 --> F
    F --> G{Visuals useful?}
    G -->|Yes| H[Pollinations + Pillow]
    G -->|No| I[Final Blog]
    H --> I

Loading

1. Input: enter a topic, target audience, and tone.
2. Route: the router decides whether current web research is needed.
3. Research: Tavily gathers evidence when required.
4. Plan: Gemini creates a structured outline.
5. Write: workers generate individual sections.
6. Merge: the reducer assembles the full Markdown article.
7. Illustrate: if visuals help, Pollinations generates art and Pillow adds exact labels.
8. Deliver: article, sources, images, and logs appear in the workspace.

🛠️ Tech Stack

Layer	Technology
Language	Python 3.10+
UI	Streamlit
Orchestration	LangGraph, LangChain
LLM	Google Gemini
Web research	Tavily Search
Image generation	Pollinations
Image processing	Pillow
Data / validation	Pydantic, Pandas

📁 Project Structure

Blog_App/
├── backend.py          # LangGraph workflow and AI generation logic
├── frontend.py         # Streamlit application UI
├── requirements.txt    # Python dependencies
├── images/             # Generated and saved blog visuals
├── *.md                # Generated blog files
└── .env                # Local API keys (never commit this file)

🚀 Quick Start

Prerequisites

• Python 3.10+
• Google AI API key (required)
• Tavily API key (optional, for web research)
• Pollinations API key (required for image generation)

Installation

# 1. Clone the repository
git clone https://github.com/<your-username>/<your-repository>.git
cd Blog_App

# 2. Create a virtual environment
python -m venv .venv

# 3. Activate it
#    Windows PowerShell:
.\.venv\Scripts\Activate.ps1
#    macOS / Linux:
source .venv/bin/activate

# 4. Install dependencies
pip install -r requirements.txt

Configuration

Create a .env file in the project root:

GOOGLE_API_KEY=your_google_api_key
TAVILY_API_KEY=your_tavily_api_key
POLLINATIONS_API_KEY=your_pollinations_api_key

Variable	Required	Used for
GOOGLE_API_KEY	✅	Gemini-powered planning and writing
TAVILY_API_KEY	➖ Optional	Web research (skipped if missing)
POLLINATIONS_API_KEY	✅ for images	AI image generation

⚠️ Never commit your .env file. Make sure .env is listed in .gitignore.

Run

streamlit run frontend.py

Open http://localhost:8501 if the browser doesn't launch automatically.

💡 Example

Input

Field	Value
Topic	How Transformers Understand Text: Tokens, Embeddings, and Self-Attention Explained
Audience	Beginners
Tone	Friendly and clear

Output: a complete Markdown article with sections, diagrams, and cited sources, saved as a .md file plus an images/ folder, ready to publish on your blog, Medium, or Dev.to.

📦 Output

• Blog Markdown files are saved in the project root.
• Generated images are saved in images/.
• The UI provides Markdown and ZIP download buttons.

🧯 Troubleshooting

ModuleNotFoundError

Make sure the virtual environment is active and dependencies are installed:

pip install -r requirements.txt

API key errors

Confirm .env is in the same directory as backend.py and contains valid keys. Restart Streamlit after changing environment variables.

Images are not generated

Check that POLLINATIONS_API_KEY is set and that the images/ directory is writable.

No research or sources in the output

If TAVILY_API_KEY is missing, the research step is skipped. Add the key and restart the app.

🗺️ Roadmap

• Multi-language blog generation (Hindi / Hinglish)
• SEO score and meta description generator
• One-click publishing to Medium / Dev.to / Hashnode
• Support for additional LLM providers
• Docker support and cloud deployment guide

🔒 Security

• Never commit .env or any API key.
• Use environment variables or a secrets manager in production.
• Review generated content and citations before publishing.

🤝 Contributing

Contributions are welcome!

1. Fork the repository
2. Create a feature branch: git checkout -b feature/your-feature
3. Commit your changes: git commit -m "Add your feature"
4. Push the branch: git push origin feature/your-feature
5. Open a Pull Request

📄 License

Distributed under the MIT License. See LICENSE for details. (Replace with the license that matches your intended use.)

🙌 Acknowledgements

Streamlit · LangGraph · LangChain · Google Gemini · Tavily · Pollinations · Pillow

---

Built by Sumit Jaiswal · If you like this project, give it a ⭐