Product / Project knowledge base

Type: Automation, Tool Description: Internal product chatbot that helps me keep track and get quicker details about funcinalities of existing complex systems. Area: Backlog, Documentation

Product / Project knowledge base

Internal product chatbot that helps me keep track and get quicker details about funcinalities of existing complex systems.

🧠 How it helps me

• Deep system understanding
  → Pulls together insights from different tools, flows, and teams to explain how things really work.
• Centralized access
  → One chat interface to query screenshots, flows, and explanations without digging through docs or threads.

🛠️ Tools Behind

• Screenshots capture → Make.com + Apify
• Flow recordings → Google AI Studio
• Retrieval-Augmented Generation (RAG) → NotebookML

⚙️ Workflow Diagram

🔍 What happens under the hood

1. Screenshots and flow recordings are captured continuously during feature development, using Make.com workflows and Google AI Studio for session recording.
2. Apify extracts context and metadata from those artifacts, storing them as structured entries.
3. NotebookML powers a RAG-based system, allowing to retrieve relevant context dynamically and generate accurate responses.
4. Through the chat interface, I can ask questions like:
   • "How does the form validation work on mobile?"
   • "What is the flow when a webhook fails for IoT gateway registration?"

📈 Use Cases

• Helping new team members understand feature flows
• Explaining complex system behavior to non-technical stakeholders
• Connecting edge cases with past support tickets or dev notes