What if people could explore my work by asking questions?

I wanted my portfolio to be more than just a static list of case studies. So I built a retrieval-augmented chatbot using Langflow, trained on my own documents.

🎯 Objective
To make my portfolio more explorable, showing how RAG can enable stakeholder self-service in a real-world context.

🧠 What it does
• Uses embeddings to chunk and store my docs
• Accepts freeform questions via chat
• Runs a semantic query to retrieve the right content
• Generates a grounded GPT response with context
• Runs directly in my Astro-based portfolio site

🛠️ Technical stack
Langflow · OpenAI API · Astro · Vector Store · RAG architecture · Prompt tuning