Skip to content

Commit 76e2b08

Browse files
committed
Updated Additional Study notes for AIP-C01 and added new Visual Study Kit page
1 parent deb1fcc commit 76e2b08

12 files changed

Lines changed: 2565 additions & 23 deletions

File tree

.vitepress/config.mts

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -297,7 +297,7 @@ export default defineConfig({
297297
{ text: 'Domain 5: Testing, Validation, and Troubleshooting', link: '/certifications/aws/aip-c01/domain-5' },
298298
{ text: 'Exam Guide', link: '/certifications/aws/aip-c01/exam-guide' },
299299
{ text: 'Cheatsheet', link: '/certifications/aws/aip-c01/cheatsheet' },
300-
{ text: 'Visual Cheatsheet', link: '/certifications/aws/aip-c01/visual-cheatsheet' },
300+
{ text: 'Visual Study Kit', link: '/certifications/aws/aip-c01/visual-cheatsheet' },
301301
{ text: 'Quick Refresher', link: '/certifications/aws/aip-c01/quick-refresher' },
302302
{ text: 'Exam Tips', link: '/certifications/aws/aip-c01/exam-tips' }
303303
]

certifications/aws/aip-c01/cheatsheet.md

Lines changed: 18 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -9,7 +9,7 @@ head:
99

1010
# AIP-C01: Cheatsheet
1111

12-
[← Overview](./index.md) · [← Exam Guide](./exam-guide.md) · [Visual Version](./visual-cheatsheet.md)
12+
[← Overview](./index.md) · [← Exam Guide](./exam-guide.md) · [Visual Study Kit](./visual-cheatsheet.md)
1313

1414
::: danger Exam Day Reference
1515
Review this page 5 minutes before the exam.
@@ -37,8 +37,15 @@ Review this page 5 minutes before the exam.
3737
| `InvokeModelWithResponseStream` | Streaming (token by token) | Low-latency UX / chat |
3838
| `InvokeAgent` | Streaming + trace | Multi-step agentic workflows |
3939

40+
**InvokeModel:** Uses **model-specific JSON** in the request body. The expected schema varies by model family.
41+
4042
**Converse API:** Use when you want a **consistent message-based interface across Bedrock models** with optional inference settings, tool config, guardrails, and prompt variables.
4143

44+
**Inference settings:**
45+
Low **temperature** = more deterministic and consistent
46+
Higher **temperature** = more creative and variable
47+
Lower **topP** = tighter token selection
48+
4249
---
4350

4451
## Agents vs. Knowledge Bases
@@ -65,6 +72,7 @@ Review this page 5 minutes before the exam.
6572
- Guardrails apply to **both inputs AND outputs**
6673
- Must be explicitly applied per API call via `guardrailIdentifier` + `guardrailVersion`
6774
- PII modes: **Redact** (mask with placeholder) vs. **Block** (reject request/response)
75+
- **Denied Topics** = business-policy blocking at the topic level, not just exact keyword blocking
6876

6977
---
7078

@@ -76,6 +84,14 @@ Review this page 5 minutes before the exam.
7684
| **Aurora PostgreSQL (pgvector)** | RDS extension | Existing PostgreSQL infrastructure |
7785
| **Amazon Kendra** | Enterprise search | NLP-powered enterprise retrieval |
7886

87+
**Retrieval concepts to remember:**
88+
- **k-NN** = nearest vectors
89+
- **Hybrid search** = semantic + keyword
90+
- **BM25** = keyword ranking
91+
- **Reranking** = re-score first-pass results
92+
- **Recall** = find more relevant chunks
93+
- **Precision** = reduce irrelevant chunks
94+
7995
---
8096

8197
## Chunking Strategies
@@ -169,4 +185,4 @@ Batch Inference = ~50% cheaper than on-demand for non-real-time high-volume jobs
169185

170186
---
171187

172-
[← Overview](./index.md) · [← Exam Guide](./exam-guide.md) · [Visual Version](./visual-cheatsheet.md)
188+
[← Overview](./index.md) · [← Exam Guide](./exam-guide.md) · [Visual Study Kit](./visual-cheatsheet.md)

certifications/aws/aip-c01/domain-1.md

Lines changed: 167 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -33,9 +33,15 @@ This is the highest-weighted domain at 31%. Focus on FM selection trade-offs, ch
3333

3434
**Context Window:**
3535
- Needed for long-document RAG, multi-turn conversations, and large prompt contexts
36-
- Claude: up to 200k tokens | Llama 3: 128k tokens | Titan: shorter context windows
3736
- Choose Claude when the scenario requires processing very long documents
3837

38+
| Model | Context Window |
39+
|---|---|
40+
| **Claude 3 (Haiku / Sonnet / Opus)** | Up to 200,000 tokens |
41+
| **Llama 3** | 128,000 tokens |
42+
| **Titan Text** | Shorter — varies by version |
43+
| **Mistral** | Varies by model size |
44+
3945
**Latency:**
4046
- Use smaller/faster models (Claude Haiku, Mistral) for real-time chat interfaces
4147
- Use larger models (Claude Sonnet/Opus) for reasoning-heavy tasks where latency is acceptable
@@ -55,6 +61,23 @@ Prefer **fine-tuning** when:
5561
- The underlying data is stable and unlikely to change often
5662
:::
5763

64+
### Titan Text vs. Titan Embeddings
65+
66+
| Model | What It Produces | Best For | Not For |
67+
|---|---|---|---|
68+
| **Amazon Titan Text** | Human-readable text | Chat responses, summaries, content generation | Semantic vector search |
69+
| **Amazon Titan Embeddings** | Numerical vectors (embeddings) | Semantic search, similarity search, retrieval pipelines | Conversational text generation |
70+
71+
**Key distinction:**
72+
- **Titan Text** is a text-generation model
73+
- **Titan Embeddings** converts text into vectors for mathematical similarity comparison
74+
- These are **not interchangeable**
75+
76+
::: tip
77+
If the question is about **semantic search**, **vector databases**, or comparing meaning using cosine similarity, think **Titan Embeddings**.
78+
If the question is about **chat, summarization, or text generation**, think **Titan Text**.
79+
:::
80+
5881
---
5982

6083
## 1.2 Prompt Engineering Strategies
@@ -77,6 +100,38 @@ Prefer **fine-tuning** when:
77100
- Low temperature (0.0–0.3): Deterministic, factual outputs — use for Q&A and summarization
78101
- High temperature (0.7–1.0): Creative, diverse outputs — use for brainstorming and creative writing
79102

103+
### Temperature vs. TopP
104+
105+
| Parameter | What It Controls | Best Use |
106+
|---|---|---|
107+
| **Temperature** | How random or deterministic token selection is | Control consistency vs creativity |
108+
| **TopP** | How much of the probability distribution is available for sampling | Control how wide the model's candidate pool is |
109+
110+
**Temperature**
111+
- Low temperature (near `0`) makes the model repeatedly choose the most likely next tokens
112+
- This produces more **deterministic, consistent, and standardized** outputs
113+
- Best for: legal contracts, policy text, structured extraction, compliance summaries, classification-style responses
114+
- The same prompt run multiple times will usually produce **nearly identical output**
115+
116+
**High temperature**
117+
- Increases variation and creativity
118+
- Best for: brainstorming, marketing copy, ideation, and multiple stylistic variants
119+
120+
**TopP**
121+
- Narrows or widens the pool of candidate tokens before sampling
122+
- Lower `topP` = tighter, safer token selection
123+
- Higher `topP` = broader selection and more variety
124+
125+
::: tip
126+
For the exam, the simplest rule is:
127+
- **Need consistency and precision?** Lower **temperature**
128+
- **Need more variety and creativity?** Raise **temperature**
129+
:::
130+
131+
::: warning Common Confusion
132+
Low temperature improves **consistency**, but it does **not guarantee correctness**. A deterministic answer can still be wrong if the prompt, context, or retrieval is wrong.
133+
:::
134+
80135
---
81136

82137
## 1.3 Data Management & RAG Pipelines
@@ -140,6 +195,48 @@ Embeddings are numerical vector representations of text. Semantically similar te
140195
| **Titan Text Embeddings v2** | AWS | General purpose, AWS-native RAG (configurable dimensions) |
141196
| **Cohere Embed** | Cohere | Multilingual retrieval, semantic search |
142197

198+
### Embedding Dimensionality Trade-Offs
199+
200+
**Embedding dimensionality** is the number of values in each vector. It is an architecture tradeoff between **semantic richness** and **efficiency**.
201+
202+
| Choice | Advantage | Cost / Trade-off | Best Fit |
203+
|---|---|---|---|
204+
| **Higher dimensions** | Captures more subtle semantic meaning and nuance | More storage, more memory use, more compute during similarity search | Complex semantic retrieval, nuanced documents, harder query understanding |
205+
| **Lower dimensions** | Faster search and lower storage cost | May lose finer semantic distinctions | Simpler retrieval workloads, tighter cost/latency constraints |
206+
207+
**How to think about it:**
208+
- Higher-dimensional vectors usually improve semantic detail, especially for complex retrieval tasks
209+
- Similarity search takes longer as dimensionality increases because more values must be compared
210+
- Lower dimensions are more efficient, but they do **not** automatically improve relevance or accuracy
211+
- The right choice depends on whether the workload prioritizes **retrieval quality** or **cost/latency efficiency**
212+
213+
::: tip
214+
For the exam, treat embedding dimensionality as a **quality vs. performance** tradeoff:
215+
- **Need richer semantic understanding?** Prefer higher dimensions
216+
- **Need lower cost, smaller storage, or faster search?** Prefer lower dimensions
217+
:::
218+
219+
### Data Quality Before Ingestion
220+
221+
Poor-quality source data leads to poor embeddings, poor retrieval, and poor model answers. If the scenario is about **validating structured data before it enters a GenAI pipeline**, think about upstream data-quality controls rather than guardrails or prompt changes.
222+
223+
**AWS Glue Data Quality** is used to:
224+
- Define explicit quality rules for pipeline data
225+
- Evaluate records against those rules before downstream processing
226+
- Produce quality scores and flag failures
227+
- Stop or quarantine bad data before it reaches embeddings, vector stores, or foundation models
228+
229+
**Examples of rules:**
230+
- `patient_id` must not be null
231+
- `admission_date` must match a valid ISO date format
232+
- Required columns must be present
233+
- Value ranges and uniqueness constraints must hold
234+
235+
::: tip
236+
Use **AWS Glue Data Quality** when the problem is **"validate and block poor-quality structured data in an ETL pipeline before model use."**
237+
Use **Guardrails** when the problem is filtering model inputs/outputs at inference time.
238+
:::
239+
143240
---
144241

145242
## 1.4 Vector Stores
@@ -163,6 +260,52 @@ Embeddings are numerical vector representations of text. Semantically similar te
163260
- Integrated with Bedrock Knowledge Bases via IAM service-linked role
164261
- Does NOT support standard OpenSearch full-text features (custom analyzers, etc.)
165262

263+
### OpenSearch as a Vector Search Engine
264+
265+
Amazon OpenSearch is one of the standard AWS answers when the scenario requires **storing embeddings** and retrieving **semantically similar items** with low latency.
266+
267+
**How to think about it:**
268+
- Store embedding vectors in the index
269+
- Run **similarity search** against those vectors
270+
- Retrieve the nearest matches for use in RAG, recommendations, or "find similar" features
271+
- Use it when the system needs to handle **large embedding collections at scale** with fast retrieval
272+
273+
**Important terminology:**
274+
- **k-nearest neighbor (k-NN)** = retrieve the `k` most similar vectors to the query vector
275+
- **Vector search / similarity search** = search by semantic closeness, not exact keyword match
276+
- **OpenSearch Serverless Vector Engine** = the Bedrock-friendly serverless option
277+
- **OpenSearch Service with the k-NN plugin** = the broader managed-service variant
278+
279+
For exam purposes, if the prompt says **store embeddings + low-latency semantic similarity search**, OpenSearch is a strong answer. If it specifically mentions **Bedrock Knowledge Bases**, default to **OpenSearch Serverless**.
280+
281+
::: tip
282+
OpenSearch is a strong answer for **semantic search**, **content recommendation**, and other "find similar items" patterns because it combines **vector similarity search** with optional **hybrid keyword search** in a managed AWS service.
283+
:::
284+
285+
### Retrieval Vocabulary You Should Know
286+
287+
| Concept | Meaning | Why It Matters |
288+
|---|---|---|
289+
| **k-NN** | Return the `k` nearest vectors to a query vector | Core concept behind semantic retrieval |
290+
| **ANN (Approximate Nearest Neighbor)** | Faster nearest-neighbor search that trades a bit of exactness for speed | Real systems use ANN for low-latency search at scale |
291+
| **HNSW** | A graph-based ANN indexing approach | Often appears when discussing high-performance vector search internals |
292+
| **Similarity Search** | Search by vector closeness instead of exact keyword match | Core retrieval behavior in RAG |
293+
| **Semantic Search** | Search by meaning, not literal wording | Explains why embeddings are useful |
294+
| **Hybrid Search** | Combine semantic search with keyword search | Often improves recall on enterprise documents |
295+
| **BM25** | Classic keyword-relevance ranking algorithm | Useful when exact terms matter alongside semantic similarity |
296+
| **Reranking** | Re-score an initial result set with a stronger model or scoring step | Improves relevance after the first retrieval pass |
297+
298+
### Retrieval Quality Trade-Offs
299+
300+
- **Recall** = how many truly relevant chunks are retrieved
301+
- **Precision** = how many retrieved chunks are actually relevant
302+
- **Higher recall** usually means retrieving more candidates, which can increase cost and noise
303+
- **Higher precision** usually means tighter filtering or reranking, which can improve answer quality
304+
305+
::: tip
306+
When the exam describes a retrieval system that misses relevant information, think about **recall**. When it retrieves too much irrelevant context, think about **precision**, metadata filtering, or reranking.
307+
:::
308+
166309
---
167310

168311
## 1.5 Compliance, Data Residency & Security
@@ -182,6 +325,29 @@ Embeddings are numerical vector representations of text. Semantically similar te
182325
Amazon Bedrock does **not** use customer prompts, completions, or training data to train the underlying foundation models. This is a built-in data privacy guarantee — any answer suggesting otherwise is wrong.
183326
:::
184327

328+
### AWS KMS and Compliance Scenarios
329+
330+
AWS KMS is the core encryption-at-rest control when the scenario emphasizes **customer control over keys**, auditability, or regulated workloads.
331+
332+
**What customer-managed KMS keys give you:**
333+
- Control over key policies and who can use the key
334+
- Control over rotation policy and lifecycle management
335+
- The ability to disable or revoke key usage if needed
336+
- Auditable key usage through AWS logging
337+
338+
**Common integrations to know:**
339+
- **Amazon S3** for document storage and RAG source data
340+
- **Amazon DynamoDB** for application state, user data, or agent metadata
341+
- **Amazon EBS** for encrypted block storage attached to EC2 workloads
342+
343+
**Audit trail:**
344+
- **AWS CloudTrail** logs KMS API activity and key usage events
345+
- This is the service to choose when the question asks for compliance-oriented auditing or evidence of encryption-key usage
346+
347+
**Exam framing:**
348+
- If the scenario says the organization needs **full control over encryption keys**, prefer **customer-managed KMS keys**
349+
- If the scenario also stresses **HIPAA**, auditability, and strict security controls, customer-managed KMS keys plus CloudTrail is a strong signal
350+
185351
---
186352

187353
## 1.6 Model Customization: Bedrock Fine-Tuning vs. SageMaker vs. RAG

0 commit comments

Comments
 (0)