View Demo • Features • Architecture • Installation • Usage

Stop searching through hundreds of pages. Ask your documents directly.

PlantMind AI is an intelligent document assistant built for industrial engineers — combining RAG, semantic search, and Google Gemini to surface critical knowledge from manuals, SOPs, and safety documents in seconds.

---

🎬 Demo

📤 Upload & Summarize	🔍 Ask Questions	🚨 Incident Analysis
:---:	:---:	:---:
📄 Upload PDF	❓ Ask Questions	⚠️ Incident Analysis
Upload any industrial PDF and get an AI-generated summary instantly	Ask natural language questions and get context-aware answers	Investigate incidents using document-grounded knowledge

---

🧠 The Problem It Solves

Traditional Approach                    PlantMind AI
─────────────────────────────────────   ──────────────────────────────────────
📂 Open maintenance manual              📄 Upload your PDF
🔍 Ctrl+F to search keywords            💬 "What is the shutdown procedure for Pump A?"
📖 Read through 200+ pages              ⚡ Get a precise, context-aware answer in seconds
📝 Copy relevant sections manually      📌 See exactly which context was used
😤 Repeat for every question            ✅ Done.

Industrial knowledge is buried across maintenance manuals, SOPs, operating procedures, and safety documentation. During maintenance windows or incident investigations, every minute of manual searching costs time — and safety.

PlantMind AI solves this by turning your static documents into an intelligent, queryable knowledge base.

---

✨ Features

📄 Document Intelligence ✅ Upload any industrial PDF ✅ Automatic text extraction with PyPDF ✅ AI-generated document summaries ✅ Smart text chunking for accuracy	🔍 RAG-Powered Search ✅ Semantic search using FAISS vector store ✅ Sentence Transformer embeddings ✅ Context-aware Q&A with Gemini 2.5 Flash ✅ Retrieved context transparency
🚨 Incident Analysis ✅ Document-grounded incident investigation ✅ SOP compliance checking ✅ Root cause assistance	🖥️ Interactive Interface ✅ Clean Streamlit web UI ✅ Real-time AI responses ✅ No backend infrastructure needed ✅ Runs fully local or on cloud

---

🏗️ Architecture

                    ┌─────────────────────┐
                    │    User Uploads PDF  │
                    └──────────┬──────────┘
                               │
                    ┌──────────▼──────────┐
                    │   PDF Text Extraction│  ← PyPDF
                    └──────────┬──────────┘
                               │
                    ┌──────────▼──────────┐
                    │    Text Chunking     │  ← Overlapping windows
                    └──────────┬──────────┘
                               │
                    ┌──────────▼──────────┐
                    │  Sentence Embeddings │  ← Sentence Transformers
                    └──────────┬──────────┘
                               │
                    ┌──────────▼──────────┐
                    │  FAISS Vector Store  │  ← Semantic Index
                    └──────┬──────────────┘
                           │
           ┌───────────────┼────────────────┐
           │               │                │
    ┌──────▼──────┐  ┌─────▼──────┐  ┌─────▼──────┐
    │ User Query  │  │  Summary   │  │  Incident  │
    └──────┬──────┘  │ Generation │  │  Analysis  │
           │         └────────────┘  └────────────┘
    ┌──────▼──────┐
    │  Semantic   │
    │  Retrieval  │
    └──────┬──────┘
           │
    ┌──────▼──────────────────┐
    │  Google Gemini 2.5 Flash │  ← Grounded in your docs
    └──────┬──────────────────┘
           │
    ┌──────▼──────┐
    │ AI Response │
    │ + Context   │
    └─────────────┘

---

🛠️ Technology Stack

Layer	Technology	Purpose
Frontend		Interactive web interface
LLM		Language understanding & generation
Embeddings		Semantic text representations
Vector Search		Fast similarity search
PDF Processing		Text extraction
Backend		Core application logic

---

📂 Project Structure

PlantMind-AI/
│
├── 📄 app.py                  # Main Streamlit application
├── 📋 requirements.txt        # Python dependencies
├── 🔐 .env.example            # Environment variable template
│
├── src/
│   ├── 🤖 llm.py              # Google Gemini integration
│   ├── 📑 pdf_processor.py    # PDF extraction & chunking
│   ├── 💬 prompts.py          # Prompt engineering templates
│   ├── 🔍 rag.py              # FAISS vector store & retrieval
│   ├── 🚨 incident.py         # Incident analysis module
│   └── 🛠️  utils.py            # Helper utilities
│
├── assets/                    # Screenshots & media
└── sample_docs/               # Example industrial PDFs

---

🚀 Installation

Prerequisites

• Python 3.10+
• A free Google AI Studio API key

Steps

1. Clone the repository

git clone https://github.com/YOUR_USERNAME/PlantMind-AI.git
cd PlantMind-AI

2. Create and activate a virtual environment

# Create
python -m venv venv

# Activate — Windows
venv\Scripts\activate

# Activate — Linux/macOS
source venv/bin/activate

3. Install dependencies

pip install -r requirements.txt

4. Configure your API key

# Create .env file
cp .env.example .env

# Add your Gemini API key inside .env
GEMINI_API_KEY=your_api_key_here

5. Run the application

streamlit run app.py

Open http://localhost:8501 — you're live! 🎉

---

📖 Usage

1. Upload PDF        →   Drag & drop any industrial manual or SOP
2. Get Summary       →   Click "Generate Summary" for an AI overview
3. Ask Questions     →   Type any question in natural language
4. View Context      →   See exactly which document sections were used
5. Analyze Incident  →   Use the incident tab for root cause analysis

💡 Example Questions You Can Ask

"What is the startup procedure for the cooling tower?"
"What are the safety precautions before entering a confined space?"
"What maintenance interval is recommended for the hydraulic pump?"
"Which SOP applies to high-pressure steam line repairs?"
"Summarize the emergency shutdown procedure."

---

🗺️ Roadmap

Status	Feature
✅	PDF upload & text extraction
✅	FAISS semantic search
✅	Gemini-powered Q&A
✅	Document summarization
✅	Incident analysis
🔄	Page-level citations
🔄	Multi-document cross-search
🔄	OCR for scanned PDFs
🔄	Persistent chat history
🔄	Export summaries to PDF
🔄	ChromaDB / cloud vector store
🔄	User authentication
🔄	Cloud deployment (Streamlit Cloud / GCP)

---

🎯 Use Cases

⚙️ Maintenance Search equipment manuals in seconds	📋 SOP Lookup Instant procedure retrieval	🔧 Troubleshooting Diagnose faults with doc context
🚨 Incident Investigation Root cause from safety docs	🛡️ Safety Compliance Verify procedures against standards	📚 Knowledge Transfer Onboard new engineers faster

---

👩‍💻 Author

Siddhant

Building AI tools that solve real engineering problems

---

📄 License

This project is open source and available under the MIT License.

---

If this project helped you, consider giving it a ⭐ — it means a lot!