AWS Services

Question 1

AWS Bedrock is used for

Answer 1

Generative AI

Question 2

I want to train my AWS Bedrock agent to be more relevant, what feature in AWS Bedrock should I use to fine-tune my agent with my own data

Answer 2

Knowledge Base (RAG) to fine tune data

also must use throughput provisioing whenever fine-tuning model

Question 3

AWS Service - Fully Managed Gen-AI assistant for your employees based on your company’s knowledge base

Answer 3

Amazon Q Business

Question 4

Can I choose the underlying FM for Amazon Q?

Answer 4

No! It is fully managed RAG service

Question 5

What service does Amazon Q Business sit on top of?

Answer 5

Bedrock

Question 6

Popular Data Sources for Amazon Q Business

Answer 6

S3, RDS, Aurora, 365, GDrive, Gmail, Slack, Sharepoint

Question 7

Popular plugins for Amazon Q Business

Answer 7

Jira, ServiceNow, ZenDesk, Salesforce

Question 8

Use cases for Amazon Q Business

Answer 8

Summaries of team meetings
Create post to advertise new role
Sending meeting invites
Create Jira Ticket

Question 9

Feature of Amazon Q Business. Create Gen AI-powered apps without coding by using natural language. Can use plugins.

Answer 9

Amazon Q Apps

Question 10

Answer questions about AWS documentation and AWS service selection. Real-time code suggestions specific to AWS that supports multiple language.

Answer 10

Amazon Q Developer

Question 11

AWS Services that can be integrated with AWS Q

Answer 11

Quicksight - ask questions and get data summaries
EC2 - provide suggestions for EC2 instance types
AWS Chatbot - Q can troubleshoot
AWS Glue - Ask questions about ETL scripts, troubleshoot

Question 12

GenAI app building playground powered by Amazon Bedrock

Answer 12

PartyRock

Question 13

NLP Service. Given several documents, can break them down into terms and classify them. Often used to understand sentiment of text

Answer 13

AWS Comprehend

Question 14

Processing types for AWS Comprehend

Answer 14

Real-time
Async analysis - batch async

Question 15

AWS Comprehend supports NER. What is NER?

Answer 15

(Named Entity Recognitition) - extract predefined entities (people, places, orgs, dates) from text

Question 16

I want to use AWS Comprehend service to alayze text for specific terms and noun-based phrase by training my data model. Does this feature exist?

Answer 16

Yes, Custom Entity Recognition

Question 17

AWS Service that translates text from one lanaguage to another

Answer 17

Translate

Question 18

What type of Algorithm does Translate use?

Answer 18

Neural Network

Question 19

Feature of Translate that allows you to trasnalte to formal vs. informal

Answer 19

Parallel data

Question 20

AWS Service that converts speech to text

Answer 20

Transcribe

Question 21

Algorithm used by Transcribe service

Answer 21

deep learning called ASR

Question 22

AWS Translate feature that allows you to capture domain specific non standard terms (technical words, acryoyms, jargon)

Answer 22

Custom Vocabularies

Question 23

AWS Translate feature that allows you to provide for context by training transcribe model on own domain specific text data to learn context associated with a given word

Answer 23

Custom Language Models

Question 24

AWS Translate feature that ML powered voice based detection capability to leverage speech cues (tone and pitch) to determine if someone is angy

Answer 24

Toxicity Detection

Question 25

AWS Service that transforms text into lifelike speech

Answer 25

Polly

Question 26

Service that allows you to find objects, people, text, scenes in images and videos using ML

Answer 26

Reckognition

Question 27

AWS Service that allows you to provide Human oversight of ML predictions in production

Answer 27

Amazon Augmented AI (Amazone A2I)

Question 28

Service that allows you to build chatbot quickly for your app supporting voice and text

Answer 28

Amazon Lex

Question 29

Services Amazon Lex can integrate with

Answer 29

AWS Lambda, Connect, Comprehend, Kendra

Question 30

Use cases for Amazon Lex

Answer 30

Allows customer to order pizzas or book a hotel

Question 31

Fully managed ML service to build personalization apps in real time

Answer 31

Amazon Personalize

Question 32

Automatically extracts text, handwriting from any scanned coument using AI/ML

Answer 32

Textract

Question 33

Use of Textract

Answer 33

Extract data from recipts

Question 34

Fully managed document search service powered by Machine Learning to extract answers from within a document (text, pdf, HTML, PowerPoint, MS Word, FAQ)

Answer 34

Kendra

Question 35

Use case of Kendra

Answer 35

Business docs all put in and can answer questions about company “where is IT located”

Question 36

Crowdsourcing marketplace to perform simple human tasks. Distributed virtual workforce (behind the scenes humans do the work)

Answer 36

Mechanical Turk

Question 37

Services Mechanical Turk integrates with

Answer 37

Integrates with Amazon A21, SageMaker Ground Truth

Question 38

Auto convert medical-related speech to text (HIPAA compliant)

Answer 38

Transcribe Medical

Question 39

Detects and return useful info from unstructured clinical text

Answer 39

Comprehend Medical

Question 40

Fully managed service to build ML models. To collect and prepare data, train ML model, deploy and monitor model.

Answer 40

SageMaker