# AI Engineering Hub > [!Note] > The discipline of building production applications on top of foundation models — > adapting, evaluating, serving and operating them, rather than training from scratch. > Distilled mainly from Chip Huyen's *AI Engineering*. Agent and RAG patterns live in > the [[6 - AI Agents and Patterns Hub]]; core model internals in > [[5 - LLM Foundations Hub]]. ## 1. The discipline - [[AI Engineering]] --- ## 2. Foundation models - [[Autoregressive Language Model]] - [[Masked Language Model]] --- ## 3. Evaluation - [[Perplexity]] - [[LLM as a Judge]] - [[AI Judge Biases]] - [[Comparative Evaluation]] - [[Evaluation-Driven Development]] - [[Data Contamination in Benchmarks]] - [[Eval-Set Sizing Heuristic]] - [[Eval Pipeline Bootstrapping]] --- ## 4. Finetuning and model adaptation - [[Parameter-Efficient Finetuning]] - [[LoRA]] - [[Quantization]] - [[Model Merging]] - [[Model Distillation]] - [[Finetuning vs RAG]] --- ## 5. Dataset engineering - [[Instruction Dataset Design]] - [[Data Synthesis for AI]] --- ## 6. Inference optimization - [[Continuous Batching]] - [[Prefill-Decode Disaggregation]] - [[Model Parallelism]] - [[Inference Goodput]] --- ## 7. Architecture and operations - [[AI Application Architecture]] - [[Model Gateway]] - [[Conversational Feedback]] --- ## 8. Anchor sources - [[AI Engineering - Chip Huyen]] --- ## Up - [[0 - Modern AI Software Engineering Hub]]