0:22 A grocery store chain has data in
0:25 several internal systems like sales,
0:27 inventory, and marketing.
0:29 Employees waste time searching these
0:31 systems for information on product
0:33 performance, stock, and campaign effectiveness.
0:35 effectiveness.
0:37 They need a central way to easily access
0:40 and understand data across these systems
0:43 for better decisions and efficiency.
0:44 Which Google Cloud offering should they
0:49 use? A Gemini for Google Workspace.
0:53 B Google agent space. C avertex AI
1:02 The correct answer is B. Google agent space.
1:14 A tech company has separate teams using
1:15 different tools for their machine
1:17 learning projects causing duplicated
1:20 work and scaling issues.
1:22 They need a central platform to manage
1:24 all their AI development, deployment,
1:27 and monitoring efficiently.
1:28 Which Google Cloud offering should they
1:32 use? Uh, cloud functions
1:35 B. Vertex AI
1:39 C. Google Agent Space D. BigQuery.
1:42 The correct answer is B. Vert.Ex AI
1:52 The correct answer is B. A Vert.Ex AI
2:03 A software company's AI chatbot
2:05 struggles to answer customer questions
2:07 about recently released features because
2:09 this information is not in its original
2:12 training data. Customers are getting
2:14 inaccurate answers, increasing support
2:16 agent workload.
2:18 The company wants the chatbot to use the
2:20 latest product documentation to give
2:23 accurate up-to-date responses without
2:26 retraining the entire model. Which
2:28 technique should they use?
2:30 uh finetuning
2:32 B prompt engineering
2:36 C retrieval augmented generation R A
2:46 the correct answer is C retrieve
2:55 a business analyst asks a generative AI
2:57 model about the quarterly revenue of a
2:59 small startup that recently entered the
3:02 market. The model confidently provides a
3:04 specific revenue figure and even
3:06 mentions a supposed press release
3:09 detailing the company's success.
3:11 However, after further investigation,
3:13 the analyst discovers that the startup
3:15 has not yet released any financial
3:17 reports and no such press release exists.
3:18 exists.
3:21 The information provided by the AI model
3:23 is entirely fabricated despite sounding plausible.
3:25 plausible.
3:27 Which type of large language model
3:30 limitation does this exemplify?
3:32 A bias,
3:36 B knowledge cut off, C data dependency,
3:50 A generative AI tool that answers
3:52 employee policy questions is providing
3:54 outdated and inaccurate information
3:56 causing confusion.
3:58 The company wants the tool to give
4:00 reliable answers based on the latest
4:02 official documents.
4:05 What should the organization do? Uh
4:07 fine-tune the underlying language model
4:09 with a broader data set of general
4:12 knowledge. B increase the temperature
4:15 setting of the language model. C.
4:17 Implement grounding techniques.
4:22 D. Reduce the token count parameter.
4:24 The correct answer is C. Implement
4:34 A sales team wants to create dynamic and
4:36 personalized video pitches for potential clients.
4:38 clients.
4:40 They receive client information in
4:42 various formats and need an AI model
4:44 that can transform this information into
4:46 engaging video content tailored to each
4:50 client's specific needs and challenges.
4:53 Which Google model should they use? A
4:56 Gemma B Gemini Cigan
4:58 Cigan Dio.
5:12 A growing retail company with fragmented
5:15 phone, email, and basic website chat
5:18 support needs a unified cloud solution.
5:20 They require integrated communication
5:22 channels, consistent customer
5:24 experiences, and scalable support that
5:27 ensures security and privacy.
5:29 Which Google Cloud offering should they
5:33 use? A Vertex AI platform,
5:36 B. Google Cloud contact center as a
5:40 service, C conversational AI,
5:43 D. A Vertex AI search. The correct
5:51 The correct answer is B. A Google Cloud
6:02 An organization is seeking to improve
6:04 how its employees access and use
6:06 internal company information scattered
6:09 across various systems.
6:10 They want to provide their knowledge
6:12 workers with tools that can understand
6:14 and use this data to enhance
6:17 productivity and decision-m.
6:19 What is a key benefit of using Google
6:22 cloud agent space in this scenario?
6:24 I agent space primarily focuses on
6:26 enhancing external customer engagement
6:29 through AI powered chat bots.
6:31 B agent space directly manages the
6:33 underlying infrastructure and hardware
6:36 required for AI model training.
6:39 C agent space allows employees to find
6:41 and use internal information more easily
6:44 by creating custom AI agents that can
6:46 access and understand data from various
6:48 enterprise sources.
6:50 D. Agent space is mainly designed for
6:52 building and deploying custom machine
6:54 learning models for predictive analytics.
6:55 analytics.
6:58 The correct answer is C. Agent space
7:00 allows employees to find and use
7:02 internal information more easily by
7:04 creating custom AI agents that can
7:06 access and understand data from various
7:09 enterprise sources.
7:11 Explanation Google Cloud agent space is
7:19 The correct answer is C. Agent Space
7:21 allows employees to find and use
7:23 internal information more easily by
7:25 creating custom AI agents that can
7:27 access and understand data from various
7:38 What is reinforcement learning? A
7:40 learning from labeled data with correct
7:42 output pairs.
7:44 Be a learning by identifying patterns in
7:47 unlabeled data. C learning through
7:49 interaction and feedback.
7:52 D learning by training on vast data to
7:55 generate new content.
7:57 The correct answer is C. Learning
8:05 A company is developing a system to
8:07 automatically categorize customer
8:08 support emails.
8:10 They have a collection of thousands of
8:13 past emails and each email has been
8:14 manually reviewed and tagged with a
8:16 category such as billing, inquiry,
8:20 technical support or feature request.
8:23 What type of data is this? A unlabeled
8:29 data, B labeled data, C structured data,
8:37 The correct answer is B.
8:47 What is the definition of a generative
8:51 AI gen AAI model? A. A physical device
8:53 that houses the hardware components of a
8:57 Gen AI system. B. A complex algorithm
8:59 trained on vast amounts of data to learn
9:02 patterns and relationships.
9:04 C. A user interface that allows users to
9:08 interact with a Gen AI system. D. A set
9:09 of rules and guidelines governing
9:12 responsible development and use of Gen AI.
9:18 The correct answer is B. A complex
9:20 algorithm trained on vast amounts of
9:22 data to learn patterns and relationships.
9:31 A video game company created a virtual
9:33 reality game with virtual characters
9:35 that can interact with users in a more
9:37 natural and intuitive way by using
9:39 gestures and facial expressions to communicate.
9:41 communicate.
9:44 What type of agent is this? A creative
9:48 agent. B a workflow agent.
9:51 C virtual assistant agent. D
9:57 The correct answer is D. A
10:05 A software company has developers who
10:07 need to write, review, debug, and
10:09 generate code from natural language
10:12 descriptions by using generative AI.
10:16 What type of agent is this? A data
10:20 analysis agent? B. A workflow agent.
10:22 C. Data agent.
10:26 DA code agent.
10:37 An advertising agency needs to quickly
10:39 create many different photorealistic
10:41 images from text descriptions for client
10:44 campaigns because traditional photooots
10:46 are too slow and costly.
10:48 The agency's goal is to generate
10:51 highquality visuals rapidly from text to
10:52 improve campaign speed and lower expenses.
10:54 expenses.
10:56 Which Google foundation model should be used?
10:58 used?
11:00 A Gemini,
11:02 B. Gemma,
11:04 C. AO,
11:18 A marketing team wants to use a
11:20 foundation model to create social media
11:23 and advertising campaigns.
11:25 They want to create written articles and
11:28 images from text. They lack deep AI
11:32 expertise and need a versatile solution.
11:34 Which Google foundation model should
11:37 they use? A. Gemma,
11:39 B. Ian,
11:41 C. Gemini, Dio.
11:59 A research scientist wants to use VO to
12:02 visualize live fluctuating data streams
12:04 on a real-time dashboard.
12:06 Why would VO be a poor choice for this
12:10 particular task? A VO generates videos
12:12 from static inputs like text or images
12:14 and cannot process or dynamically
12:16 visualize live data.
12:19 Be vio may lack specific scientific
12:21 visualization styles needed for accurate
12:24 data representation.
12:26 C. A video use for this scenario would
12:27 require too many computational resources.
12:29 resources.
12:32 DO is designed for short form video not
12:35 continuous longduration live data displays.
12:40 The correct answer is a VO generates
12:43 videos from static inputs like text or
12:45 images and cannot process or dynamically
12:54 A company sells custom-designed phone
12:57 cases on their website. How should Ian
12:59 be effectively used?
13:02 A to generate realistic images of phone
13:04 cases on devices from text descriptions
13:06 of designs.
13:08 B. To analyze customer feedback to
13:10 identify popular phone case design trends.
13:12 trends.
13:14 C to predict demand for different phone
13:18 case designs based on sales data. D. To
13:20 transcribe customer audio feedback on
13:23 prototypes of phone case designs.
13:27 The correct answer is A.
13:30 The correct answer is A. To generate
13:32 realistic images of phone cases on
13:33 devices from text descriptions of designs.
13:42 A company lacks extensive in-house
13:45 machine learning and AI expertise.
13:49 How does Google Cloud democratize AI? Uh
13:51 by offering exclusive access to its most
13:53 powerful AI models for high-spending clients.
13:55 clients.
13:57 By providing fully automated AI
14:00 solutions requiring no technical user input.
14:01 input.
14:04 C. By providing a comprehensive AI
14:06 platform with lowode/noodeode tools,
14:10 pre-trained models and easytouse APIs.
14:13 D by offering free custom AI solution
14:15 development and deployment for all
14:24 The correct answer is C. by providing a
14:25 comprehensive AI platform with
14:29 lowode/nood tools, pre-trained models,
14:38 A marketing team is brainstorming
14:40 creative ideas for a new social media campaign.
14:42 campaign.
14:43 They want to quickly generate various
14:46 taglines and initial post drafts to
14:48 explore different angles and messaging options.
14:49 options.
14:51 They need a Google Cloud pre-built
14:53 generative AI tool without additional
14:55 setup that allows for rapid content
14:58 creation and experimentation.
15:01 What should the team do? Uh, use
15:03 Notebook LM to upload past successful
15:05 campaigns and analyze their key elements.
15:06 elements.
15:09 B. Use Gemini for Google Workspace
15:11 within Google Slides to prepare a
15:12 presentation outlining their social
15:15 media campaign strategy.
15:18 C. Use the Gemini app to create taglines
15:20 and social media post drafts that align
15:22 with their campaign goals and target audience.
15:23 audience.
15:26 D. Create a custom gem in Gemini
15:28 Advanced with specific brand guidelines
15:36 Correct answer is C. Use the Gemini app
15:38 to create taglines and social media post
15:40 drafts that align with their campaign
15:47 A data science team needs a centralized
15:49 and organized location to store its
15:51 various model versions, track their
15:54 metadata, and easily deploy them to the
15:56 respective applications.
15:58 What Google Cloud service should they
16:01 use? A cloud storage,
16:04 B. Model registry,
16:06 C. BigQuery,
