LogWise is a fast, lightweight and intelligent log analyzer built to help developers, DevOps engineers, and SREs gain quick insights from long and messy server log files. Powered by LLMs like Mistral (via Ollama), it summarizes, identifies errors, and provides useful diagnostics — instantly.
Developers often struggle with:
- Endless
.logfiles filled with stack traces, warnings, timestamps, and cryptic messages - Manually scanning logs to debug production issues
- Missing key events hidden deep in noisy outputs
Logs are essential — but often overwhelming.
LogWise turns raw logs into natural language summaries, identifying:
- 🔧 Root causes of failures
⚠️ Warnings and timeouts- 🧠 Insights and optimization tips
- 🔍 AI Summarization of complex logs using local LLMs
- ⚡ Blazingly fast file upload & analysis
- 🧠 Context-aware analysis (not just pattern matching)
- 🌙 Fully offline with Ollama
- 🎨 Beautiful UI built with Tailwind CSS
- 🔐 No data is sent to cloud APIs
| Layer | Tech Used | Description |
|---|---|---|
| 🧠 AI Engine | Ollama + Mistral | Local LLM used for intelligent log parsing |
| 🚀 Backend | FastAPI (Python) | REST API to handle file upload and analysis |
| 🎨 Frontend | React + Vite + Tailwind CSS | Modern UI, responsive and clean |
| 🐳 DevOps | Docker + Docker Compose | Containerized app environment |
| 🧪 Tools | Axios, FormData | Frontend-backend communication |
logwise/
├── backend/ # FastAPI backend
│ ├── main.py
│ ├── routes.py
│ ├── utils.py
│ └── requirements.txt
│
├── frontend/ # React + Vite frontend
│ ├── public/
│ ├── src/
│ │ ├── pages/
│ │ ├── components/
│ │ └── main.jsx
│ └── index.html
│
├── docker-compose.yml # Fullstack dev setup
├── .gitignore # Clean repo
└── README.md # This fileUpload this log file:
2024-07-01 12:02:15 WARNING Disk usage exceeds 85%
2024-07-01 12:02:18 ERROR Timeout occurred in UserService at line 45
2024-07-01 12:02:20 INFO Reconnecting to database...→ You'll get an AI summary like:
1. Disk space is critically high.
2. A timeout error occurred in UserService.
3. A recovery attempt to reconnect was initiated.# Start both frontend and backend
docker-compose up --buildThen open: http://localhost:5173
✅ Make sure you have Ollama installed locally and model pulled:
ollama pull mistral