RAG Pipeline Logic

Ingestion

• Script: backend/ingest.py
• Process:

1. Scans docs/. 2. Cleans MDX (removes frontmatter/imports). 3. Chunks text (1000 chars, 100 overlap). 4. Embeds using models/text-embedding-004. 5. Upserts to Qdrant collection physicalaibook.

• Run: python backend/ingest.py

Vector Search (Qdrant)

• Client: qdrant-client
• Collection: physicalaibook
• Vector Size: 768 (Gecko-004)
• Similarity: Cosine

Prompt Engineering

• File: backend/utils/helpers.py.
• RAG Prompt: Constructs a prompt containing retrieved context chunks.
• Personalization: backend/personalization.py creates system instructions based on softwarebackground and hardwarebackground of the user.

Agentic Flow

We use a custom Agent class (backend/agents.py) that wraps the LLM calls, allowing for future expansion into multi-agent workflows.