1. Project Overview

What I Built: An AI-powered DevOps Knowledge Base Auto-Builder using Notion MCP, Python, and Groq's LLM. The system takes raw, unstructured incident reports — the kind engineers write in Slack or a notes app after a stressful outage — and automatically extracts structured knowledge, then creates a formatted Notion database entry with zero manual work.

The Problem It Solves: After incidents, engineers are exhausted. Documentation gets skipped. Knowledge dies. This tool means you dump your raw notes immediately and your KB builds itself — root cause, fix, prevention steps, tags, severity, all populated automatically.

Stack Used:

• Python 3.10 (WSL on Windows)
• Groq API (Llama 3.3 70B) for AI extraction
• Notion API for database creation
• Notion MCP integration for workspace access
• Aider for AI-assisted coding

2. Environment Setup

Machine Constraints

This was built on a Windows 32-bit machine without a Claude subscription. All work was done inside WSL (Windows Subsystem for Linux) using free-tier tools — Groq's free API and Notion's free plan. No paid subscriptions required.

Installing Dependencies

First, I fixed the PATH issue so tools installed via pip were accessible:

export PATH="$HOME/.local/bin:$PATH"
echo 'export PATH="$HOME/.local/bin:$PATH"' >> ~/.bashrc
source ~/.bashrc

Then installed Aider and the required Python packages:

pip install aider-chat --user
pip install requests groq python-dotenv --user

Screenshot 1: Take a screenshot of your WSL terminal showing the successful aider --version output (showing aider 0.86.2)

Project Folder Setup