AI Engineering · 400-level
AI System Design
AIE-4033 creditselectivebadge: architectprereqs: AIE-301, AIE-401
Start this course — free
Earn the architect credential
What's inside
Sections & lessons
01
Gateway/proxy pattern
- Architecture tradeoffs behind a real AI pipelineconcept45 min
02
Caching strategies at the system level
- Where a cache belongs in a multi-node pipelineconcept60 min
03
Load simulation
- Steady, spike, and adversarial-burst traffic against your own designconcept60 min
04
Observability: traces, cost attribution
- Naming the exact failure mode before it shipsconcept60 min
05
Assemble a full pipeline
- Ingestion → embed → retrieve → rerank → generate → guardrailconcept90 min
06
Quiz: architecture-tradeoff judgment
- Checkpoint on where a design is most likely to break firstconcept30 min
07
Lab + eval-gate: find the breaking node
- Defend a capacity/cost report against a simulated traffic ramp (Architect badge)concept120 min
Learn it from the inside
This module's playgrounds
Vocabulary
Key concepts in this course
Vector Database Reranking Retrieval-Augmented Generation (RAG) Latency Percentiles (p50/p95/p99) Prompt Caching
Compare related approaches
Optional · watch & try
Go deeper, elsewhere
Hand-picked public explainers and open tools — always optional, never required, never graded.
A Visual Tour of Modern LLM ArchitecturesCompares architecture choices across current model families LLM Architecture Explained | From Neural Networks to ResponsesEnd-to-end system view from input to served response AI Agent Architecture Explained: LLMs, Context & Tool ExecutionSystem-design lens specifically for agentic applications Artificial Analysis (model comparison)Compare quality/speed/price across providers when designing a routing/gateway layer
This module ends in a gate you can fail.
That's what makes passing it mean something. Take the DSAT, get placed, and start earning.