File size: 1,624 Bytes
17a77b6
 
 
 
 
 
 
 
 
 
 
35cccae
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1ad37d2
35cccae
1ad37d2
35cccae
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1ad37d2
35cccae
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
---
title: WatchTowerAI - Log Analyzer
emoji: 🛰️
colorFrom: purple
colorTo: pink
sdk: streamlit
sdk_version: "1.32.0"
app_file: streamlit_app.py
pinned: false
---

#  WatchTowerAI – Log Classification & Automated Runbook Generator

A production-grade AI tool to analyze logs in real-time, classify them using a fine-tuned Flan-T5 model, generate markdown runbooks, and automatically alert your team via Slack — all deployed on Hugging Face Spaces.

##  Features

-  **LLM-powered log classification** (INFO, WARNING, ERROR, CRITICAL, SECURITY)
-  **Runbook generation** using Flan-T5
-  **Upload runbooks to AWS S3**
-  **Send real-time Slack alerts**
-  Hybrid fallback using rule-based logic if model is uncertain
-  Streamlit Web UI + `.log` file upload support

##  Tech Stack

- LLM: Fine-tuned [Flan-T5](https://huggingface.co/google/flan-t5)
- Frontend: Streamlit
- Cloud: AWS S3
- Alerts: Slack Webhook
- Deployment: Hugging Face Spaces

##  Screenshot

![screenshot](https://your-screenshot-link)

##  How to Use

1. Paste or upload a `.log` file
2. Click “Classify + Generate Runbook”
3. Download the markdown or view it in S3
4. Slack alerts will be sent automatically

##  Directory

- `streamlit_app.py` – UI + logic
- `upload_to_s3.py` – S3 upload function
- `notify_slack.py` – Slack integration
- `runbooks/` – Markdown output folder
- `codementor-flan/` – Your fine-tuned model folder

##  AI Model

Fine-tuned on domain-specific QA + system logs using Flan-T5 for robust multi-class log categorization and response generation.

---

 Made by [Chetan](https://github.com/chetan10510)