Google Professional-Agentic-Architect actual dump : Google Cloud Certified - Professional Agentic Architect

You are configuring Agent Platform Memory Bank for a wealth management advisor agent deployed on Gemini Enterprise Agent Platform.
During the initial evaluation, you observe that Memory Bank creates duplicate and conflicting memory entries when users update their investment risk tolerance. You need to configure Memory Bank to properly consolidate updated user preferences over time, resolve conflicting information, and restrict memory extraction strictly to relevant personal information and user preferences.
What should you do?
Choose two

• A. Specify a consistent user identity scope during memory generation and retrieval.
• B. Add user risk tolerance values into the system instruction prompt as immutable string constants during session creation.
• C. Disable identity scoping on the Memory Bank instance to do a similarity search for user information during memory consolidation.
• D. Configure memory topic customization in Memory Bank to enable the USER_PERSONAL_INFO and USER-PREFERENCES managed topics.
• E. Set an in-memory FIFO eviction policy on the managed session service to automatically purge older session event logs when user profile updates occur.

You are developing a customer support agent in Gemini Enterprise to analyze incoming customer incident reports. These reports include recorded audio calls in MP3 and WAV formats and long video-screen captures lasting up to 45 minutes. When you call the Gemini API with a raw Base64 payload, the request fails with a
413 error: size too large. You need to enable the agent workflow to process, analyze, and summarize these large audio and video files natively using Gemini models, without hitting payload restrictions or requiring preconversion steps.
What should you do?

• A. Convert audio files into 16-bit PCM WAV chunks, and embed them using the text-embeddings API.
• B. Upload the audio and video files to a Cloud Storage bucket, and pass the Cloud Storage URIs in the request content to the Gemini API.
• C. Store compressed video and audio files in a Cloud SQL database, and configure a Gemini Enterprise database connector to stream binary objects.
• D. Transcode audio and video files into Base64 chunk strings, and send them sequentially through Pub
  /Sub messages to the Gemini API endpoint.

You are building a multiagent workflow in Antigravity in which a primary coordinator agent delegates code editing and testing tasks to concurrent subagents. During initial execution, the subagents edit files in the shared workspace simultaneously, causing write conflicts and filling the main chat history with verbose build logs. You want to eliminate file modification conflicts and prevent context rot in the primary coordinator agent.
What should you do?

• A. Instantiate subagents programmatically with a canary batch cap of three concurrent workers, and store context logs in BigQuery.
• B. Configure subagents to use inherited workspace contexts for code edits, and deploy a PreToolUse hook to upload context logs to a Cloud Storage bucket. Use the stored context for additional prompts.
• C. Define PreToolUse hooks to restrict subagent file write tools, and route build traces to Cloud Logging.Retrieve the stored logs for additional prompts.
• D. Configure subagents to use branched workspace isolation for code edits, and offload build execution to isolated worker context sessions.

You are setting up an automated unit testing pipeline in which Antigravity generates and executes test scripts for third-party Python packages. Your organization's security requirements state that dynamic test suites generated by the LLM must never access local host credentials, private developer files, or unapproved internal network services. You need to configure a secure execution runtime for the coding agent.
What should you do?

• A. Assign a Google Cloud service account with restricted, least-privilege IAM roles to the pipeline execution host.
• B. Execute agent-generated test scripts inside isolated, containerized execution sandboxes.
• C. Configure a .antigravityrules markdown file in the repository instructing the agent to never generate code that reads host files or connects to external domains.
• D. Configure a VPC Service Controls perimeter around the pipeline's execution subnet.

You manage a support agent that uses RAG to answer questions based on a document corpus in RAG Engine.
Each query retrieves the top 50 chunks. While reviewing a set of failing questions, you determine that the chunk containing the answer is among the 50 returned but is ranked near the bottom, and the agent answers from a higher-ranked chunk instead.
You want the answering chunk to rank at the top without re-indexing the corpus. The retrieval step has 200 milliseconds of headroom.
What should you do?

• A. Set a vector_distance_threshold on the RAG Engine retrieval configuration to return more relevant chunks.
• B. Add the Agent Platform ranking API as the Agent Search reranker to reorder the chunks.
• C. Update the RAG Engine deployment configuration from the Spanner mode Basic tier to the Scaled tier.
• D. Add a reranker that calls a Gemini model to score each retrieved chunk, and reorder the chunks based on the user's input.