Chapter 3 of 5 · AWS AIF-C01
Applications of Foundation Models
Model selection, RAG, prompt engineering, fine tuning and how to evaluate FM output.
- 7topics
- 28 minfast track
- 4diagrams and exhibits
- 28%of the exam
Topics in this chapter
Take them in order, or jump to the one you need.
- 1FM Selection and Inference SettingsPick a foundation model for an application using cost, modality, latency and customisation needs, and tune output with…4 min
- 2Retrieval Augmented GenerationUnderstand how RAG grounds a foundation model in your own data using embeddings and vector stores, and how Amazon Bedrock…Diagram4 min
- 3Customisation Trade Offs and AgentsCompare prompt engineering, RAG, fine tuning and training from scratch by cost and effort, and see how Amazon Bedrock Agents…Diagram4 min
- 4Prompt Engineering TechniquesLearn prompt engineering concepts and techniques such as zero shot, few shot, chain of thought, negative prompts and prompt…4 min
- 5Prompt Best Practices and RisksApply prompt best practices and understand prompt risks such as injection, hijacking, jailbreaking, poisoning and exposure, with…Diagram4 min
- 6FM Training and Fine TuningUnderstand pre training, fine tuning, continued pre training, distillation and RLHF for foundation models, plus how to prepare…Diagram4 min
- 7Evaluating FM PerformanceEvaluate foundation models with human review, benchmark datasets and metrics such as ROUGE, BLEU and BERTScore, and link results…4 min
About this chapter
What this chapter covers
This chapter is about using foundation models (FMs) in real applications. It covers how to choose and set up a model, how to give it your data, how to write good prompts, how to customise it, and how to judge its results.
It follows the third domain of the AIF-C01 outline, which has four topics.
This is the largest part of the exam, so expect many scenario questions that ask you to pick the best approach.
How the topics fit together
The pages follow the order in which a team builds an FM application.
- Design considerations cover model choice, inference settings, RAG, customisation cost and agents.
- Prompt engineering improves output by changing the input, with risks to know.
- Training and fine tuning change the model itself when prompts and RAG are not enough.
- Evaluation shows whether the result is good and whether it meets the business goal.
The usual exam logic is to choose the cheapest approach that works: prompt engineering first, then RAG, then fine tuning.
How to use these pages
Each topic page has the same shape, so you can skim for what you need.
- What it is: the idea in plain words.
- How it works: the mechanism, with diagrams and tables.
- What to remember for the exam: the facts to memorise.
- Common mistakes: the classic traps.
- Check yourself: three questions to test your understanding.
Build a small comparison table in your own notes for prompting, RAG, fine tuning and continued pre training, because many questions ask you to choose between them.
Ready for more practice?
The full AWS AIF-C01 question bank has 849 questions, with timed practice exams and an explanation for every answer.
This study guide is independent and is not affiliated with, endorsed by, or sponsored by AWS. Vendor names, certification names and exam codes are trademarks of their respective owners and are used for identification only.