Skip to content
2 min read · 381 words

Module 5: Advanced AI Engineering

**Advanced**  ·  **~30 min**  ·  **7 lessons**

Overview

Module 3 covered the RAG-relevant services in isolation; this module traces the whole pipeline end to end and fills in everything around it — embeddings, ChromaDB internals, the LangGraph concepts behind Corrective RAG, and three areas this backend deliberately (or accidentally) doesn't implement: background tasks, real conversation memory, and OpenAI-native tool calling. Every "not implemented" lesson here tells you exactly what's missing and what adding it would take.

What You'll Learn

  • The six-stage RAG pipeline: upload → chunk → embed → store → retrieve → evaluate-then-generate-or-refuse
  • The real embedding model and dimensionality (1536, corrected from an earlier draft's 384)
  • ChromaDB's isolation model — and a real orphaned-vector resource leak found in the source
  • Why Corrective RAG uses a graph specifically, not just an if statement
  • Why chat_with_agent has no real conversation memory despite persisting full history

Prerequisites

Lessons in This Module

#LessonWhat it covers
14RAG PipelineThe six-stage pipeline end to end
15EmbeddingsThe real model, dimensionality, cost
16Vector Databases & ChromaDBCollections, isolation, the orphaned-vector bug
17LangGraph & Corrective RAGWhy a graph, specifically, for this one pipeline
18Background TasksConfirmed absent — what adding them would take
19Prompt Engineering & Tool CallingAll six prompts; why this isn't OpenAI-native tool calling
20Memory & Context ManagementWhy chat has no real memory

Learning Outcomes

By the end of this module you can:

  • Trace a single uploaded document through every table it touches, SQL and ChromaDB both
  • Explain why ingestion and query embeddings must come from the same model
  • Explain what "Corrective" adds to plain RAG and what it costs (one extra LLM call)
  • Explain precisely why a chat conversation has no memory today, and what the minimal fix looks like

Start Module → Back to Module 4