Generative AI Leader Hard Questions — 25 Difficult Questions

Question 1

A legal team wants to use GenAI to review contracts and highlight risky clauses. They need the AI to consistently follow a specific classification taxonomy. The team has a small set of labeled examples (500 contracts). Which approach yields the BEST accuracy for this use case?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Fine-tune a base model using the labeled examples in Vertex AI. Fine-tuning a base model on the 500 labeled contracts teaches the model the specific classification taxonomy directly in its weights, producing consistent, high-accuracy outputs that follow the exact label set. With a small but well-labeled dataset, supervised fine-tuning is the most reliable way to enforce a fixed taxonomy rather than relying on prompt instructions.

Answer

Prompt engineer a large foundation model with few-shot examples in Vertex AI Studio

Answer

Use the RAG Engine to retrieve similar clauses and ask the model to classify

Answer

Use a larger foundation model without fine-tuning and rely on its pre-trained knowledge

Question 2

A data scientist is using Vertex AI Model Registry to manage multiple versions of a custom text classification model. They need to ensure that only the version that passes all evaluation metrics can be deployed to a Vertex AI Endpoint for online predictions. What deployment strategy should they use?

Accepted Answer

Set up continuous evaluation with model monitoring to auto-promote versions that meet thresholds. Vertex AI Model Registry supports continuous evaluation pipelines that automatically promote a model version to a 'default' or 'deployed' status only when it meets predefined evaluation thresholds (e.g., accuracy, F1 score). This ensures that only the version passing all metrics is eligible for deployment to a Vertex AI Endpoint, enabling automated, gated rollouts without manual intervention.

Answer

Use A/B testing between versions and manually select the best performer

Answer

Manually track version IDs and deploy the latest version

Answer

Deploy all versions to a single endpoint and route traffic manually

Question 3

A research lab is using Vertex AI to generate high-resolution medical images (2560x1920) of cell structures using Imagen. They have fine-tuned the model on their own microscope images. The generated images are sharp but often contain repeating patterns (e.g., identical cell arrangements) that are not biologically plausible. The team suspects the model is overfitting to spatial patterns in the training data. They have already tried increasing the training dataset size and augmenting it with rotations and flips. What additional technique should they try within Vertex AI?

Accepted Answer

Add regularization techniques such as dropout layers or data augmentation that randomly crops and blends patches.. The repeating patterns indicate the model is memorizing spatial arrangements rather than learning generalizable features. Adding regularization like dropout layers or data augmentation that randomly crops and blends patches (e.g., CutMix or MixUp) directly reduces overfitting by forcing the model to focus on local, biologically plausible details rather than memorizing entire image layouts. Vertex AI's training pipelines support custom augmentation strategies, making this a practical and targeted fix.

Answer

Switch to a different foundation model like Stable Diffusion.

Answer

Use a larger batch size during fine-tuning.

Answer

Further increase the resolution of training images to 5120x3840.

Question 4

A regulated industry client requires that all AI model predictions be logged with a traceable audit trail, including the model version, input data, and output, for compliance with internal policies. Which Vertex AI feature should they enable?

Accepted Answer

Vertex AI Model Monitoring (with prediction logging). Vertex AI Model Monitoring with prediction logging captures model version, input data, and output for every prediction request, storing them in BigQuery or Cloud Logging to create a traceable audit trail. This directly meets compliance requirements for regulated industries by providing immutable logs that can be queried and audited. Other Vertex AI features like Explainable AI or Feature Store do not offer the same comprehensive logging and version tracking needed for audit trails.

Answer

Vertex AI Explainable AI

Answer

VPC Service Controls

Answer

Vertex AI Feature Store

Question 5

A machine learning engineer needs to deploy a Gemini model on an Android device for offline inference (no internet connection) to provide real-time suggestions. Which Gemini model variant is MOST appropriate?

Accepted Answer

Gemini Nano. Gemini Nano is specifically designed for on-device inference with small memory footprint. Ultra, Pro, and Flash are cloud-only models that require internet connectivity.

Answer

Gemini Ultra

Answer

Gemini Pro

Answer

Gemini Flash

Question 6

A manufacturing company wants to use GenAI to generate maintenance reports from sensor data. They need structured output (JSON) for downstream systems, and they want to reduce token costs. Which THREE strategies should they use?

Accepted Answer

Use batch API requests for multiple sensor readings. Structured output ensures JSON format; batch requests reduce cost; the smallest suitable model minimizes token usage. Few-shot adds tokens; caching may not help for diverse sensor data.

Answer

Use the largest available model to ensure accuracy

Answer

Include multiple few-shot examples of JSON in every prompt

Question 7

A global corporation with 50,000 employees has seen rapid adoption of GenAI across marketing, product, and engineering teams. Each team selected its own models and cloud accounts, resulting in fragmented governance, unexpected costs, and varying output quality. The CFO demands a unified strategy to control costs and ensure consistency. The Chief AI Officer proposes several solutions. Which course of action best balances control with innovation?

Accepted Answer

Establish a GenAI Center of Excellence (CoE) that provides approved models, shared APIs, and best practices, while allowing team-specific customizations. A GenAI Center of Excellence (CoE) provides centralized governance through approved models and shared APIs, enabling cost control and quality consistency while preserving team-level flexibility for innovation. This balances the CFO's need for unified strategy with the CAIO's goal of avoiding rigid mandates that stifle experimentation.

Answer

Migrate all GenAI workloads to a single on-premises server to reduce cloud costs

Answer

Mandate all teams use a single model (e.g., Gemini) via a centralized Vertex AI endpoint with usage quotas

Answer

Allow teams to continue using their own models but require them to submit monthly cost reports

Question 8

A company has deployed a GenAI-powered report generation system using Vertex AI. They notice that the cost is higher than expected. Investigation shows that many requests include very long prompts with repetitive boilerplate text. Which cost optimization strategy is MOST effective?

Accepted Answer

Enable context caching for repeated prompt prefixes. Caching repeated prompt prefixes can significantly reduce token usage and cost. Batch requests help with throughput but not with per-request token savings. Reducing model size may hurt quality. Ignoring is not a strategy.

Answer

Increase the batch size for batch requests

Answer

Switch to a smaller model size

Answer

Ignore the cost increase as it will stabilize

Question 9

A legal firm wants to automate contract analysis to extract key clauses and risks. They have 10,000 contracts in PDF format. The solution must handle varying layouts and be cost-effective. Which approach is BEST?

Accepted Answer

Use Document AI to convert PDFs to structured text, then use Vertex AI with a prompt that specifies clauses to extract. Using Document AI to parse PDFs into text, then Vertex AI with a structured prompt for clause extraction combines robust document understanding with flexible GenAI. Fine-tuning on 100 contracts is insufficient for layout variation. Agent Builder is overkill. Direct Gemini API on raw PDFs loses document structure.

Answer

Build a retrieval-augmented generation (RAG) system in Vertex AI Agent Builder

Answer

Use the Gemini API directly with the raw PDF files as input

Answer

Fine-tune a Gemini model on 100 annotated contracts and run inference on all contracts

Question 10

Which THREE are valid methods to reduce bias in generative AI outputs?

Accepted Answer

Using a more diverse training dataset. Option C is correct because a more diverse training dataset reduces representational bias by exposing the model to a wider range of demographics, cultures, languages, and viewpoints, so the learned distribution is less skewed toward a dominant group. Option D is correct because safety filters (e.g., content moderation classifiers, toxicity detectors, or bias guardrails applied to prompts and outputs) can detect and block biased or harmful generations, mitigating bias at inference time. Option E is correct because prompt engineering that explicitly instructs the model to be fair, neutral, or inclusive (e.g., system prompts specifying balanced representation) steers generation toward less biased outputs without retraining. Option A is not correct because restricting prompts to English narrows the linguistic and cultural scope, which can amplify bias rather than reduce it. Option B is not correct because increasing model size alone does not remove bias and can even amplify biases present in the training data.

Answer

Using only English prompts

Answer

Increasing model size

Question 11

A research team is training a large language model from scratch using TPUs on Google Cloud. Which storage solution provides the highest throughput for training data?

Accepted Answer

Cloud Storage. Cloud Storage provides the highest throughput for training data because it is designed for high-bandwidth, parallel access from TPU pods via the Google Cloud Storage FUSE or gRPC-based data loading. TPUs benefit from Cloud Storage's ability to serve data at hundreds of GB/s when using the `tf.data` service with `tf.io.gfile` or the `gcloud storage` API, avoiding the I/O bottlenecks of block storage. Persistent Disk and Filestore have lower aggregate throughput limits and are not optimized for the distributed, streaming read patterns typical of large-scale training.

Answer

Persistent Disk

Answer

Cloud Filestore

Answer

Cloud Spanner

Question 12

A company has been using an on-premises ML infrastructure for generative AI and wants to migrate to Google Cloud. They have a pipeline that fine-tunes a large language model weekly using a proprietary dataset. The migration must minimize downtime and data transfer costs. Which approach best addresses these requirements?

Accepted Answer

Use Vertex AI Pipelines to orchestrate the fine-tuning process, and use Vertex AI Managed Datasets to incrementally sync new data with BigQuery as the source.. Vertex AI Pipelines provides a managed, serverless orchestration service that can run the weekly fine-tuning workflow with minimal operational overhead, while Vertex AI Managed Datasets can incrementally sync new data from BigQuery, reducing data transfer costs by avoiding full dataset copies. This combination minimizes downtime because the pipeline can be triggered on a schedule without manual intervention, and incremental syncs avoid re-transferring the entire proprietary dataset each week.

Answer

Use AutoML to train a new model directly from the dataset without fine-tuning.

Answer

Deploy the existing pipeline on a Google Kubernetes Engine cluster and use Google Cloud Filestore for shared storage.

Answer

Use Cloud Storage Transfer Service to move all data to Cloud Storage, then set up a Vertex AI custom training job to run the fine-tuning.

Question 13

An organization wants to ensure that when their employees use the Gemini API via Vertex AI, the grounding searches are restricted to internal company knowledge bases rather than the public web. Which feature should they enable?

Accepted Answer

Vertex AI Search (with private data indexing). Vertex AI Search allows organizations to index private, internal data sources (e.g., documents, databases) and use them as the grounding source for Gemini API queries. By enabling this feature, grounding searches are restricted to the indexed private knowledge base, ensuring no public web results are used. This directly meets the requirement to keep grounding internal.

Answer

Enterprise Data Governance on Vertex AI

Answer

VPC Service Controls

Answer

Google Search Grounding

Question 14

A financial services firm is deploying a generative AI chatbot for customer inquiries. They have strict compliance requirements: all conversations must be auditable and the model must not use customer data for training. Which Google Cloud offering should they choose?

Accepted Answer

Vertex AI Agent Builder with data governance controls. Vertex AI Agent Builder is correct because it provides built-in data governance controls that prevent customer data from being used for model training, while also supporting full auditability through integration with Cloud Audit Logs and Cloud Logging. This directly addresses the firm's compliance requirements for auditable conversations and data privacy.

Answer

Private Google Access for on-premises connectivity

Answer

Dialogflow CX with Cloud Logging

Answer

Cloud AI Platform Pipelines

Question 15

A company wants to generate a video from a text description using Google Cloud. Which service is designed for this?

Accepted Answer

Veo. Veo is Google Cloud's generative AI model specifically designed for creating high-quality videos from text or image prompts. It leverages advanced diffusion and transformer architectures to generate coherent video sequences, making it the correct choice for text-to-video generation.

Answer

Codey

Answer

Chirp

Answer

Imagen

Question 16

An e-commerce company uses a generative AI model to generate product descriptions. They observe that descriptions for high-end products use more sophisticated language compared to budget products, potentially reinforcing class stereotypes. What is the most likely cause, and what should they do to mitigate it?

Accepted Answer

The training data reflects real-world associations; fine-tune with a balanced dataset that includes diverse product descriptions across price ranges. The bias stems from training data correlations. Fine-tuning on balanced, diverse data can reduce stereotypical associations. The other options either do not address the root cause or are less effective.

Answer

The safety filters are too aggressive; relax them

Answer

The temperature parameter is set too low; increase it to introduce more randomness

Answer

The model architecture is biased; switch to a different base model

Question 17

An organization is using Vertex AI to fine-tune a large language model. They notice training is taking longer than expected and cost is increasing. Which action is most likely to reduce training time and cost without significantly impacting model quality?

Accepted Answer

Enable mixed-precision training (bfloat16). Mixed-precision training with bfloat16 reduces memory usage and accelerates computation by using half the bits of standard float32, which directly decreases training time and cost on TPUs and modern GPUs. Vertex AI supports bfloat16 natively on TPU v3+ and A100 GPUs, and for many large language models, this precision preserves model quality because bfloat16 retains the same exponent range as float32, avoiding underflow issues common with float16.

Answer

Increase the number of training steps

Answer

Increase the batch size

Answer

Use a higher learning rate

Question 18

A company deploys a Gemini model on Vertex AI for a healthcare application. They need to ensure that the model does not generate medical advice and that responses are grounded in trusted medical sources. Which combination of safety measures should they implement?

Accepted Answer

Enable safety filters and use Vertex AI Grounding with a labeled medical dataset. It combines two essential safety layers: safety filters block harmful content (including medical advice), and Vertex AI Grounding anchors responses to a labeled medical dataset, ensuring factual accuracy and compliance with healthcare regulations. This dual approach prevents the model from generating unverified or dangerous medical information while maintaining relevance to trusted sources.

Answer

Use Vertex AI Grounding with a public dataset and disable safety filters

Answer

Enable safety filters only, without grounding

Answer

Fine-tune the model on a curated medical dataset and disable safety filters for faster responses

Question 19

A cloud architect is designing a generative AI pipeline that must comply with the EU AI Act for high-risk AI systems. Which of the following is a mandatory requirement under the Act?

Accepted Answer

The system must undergo a conformity assessment before deployment. Under the EU AI Act, high-risk AI systems must undergo a conformity assessment before deployment to ensure compliance with requirements such as risk management, data governance, and transparency. This is a mandatory procedural step, not a performance metric or specific reasoning technique. The assessment may involve self-evaluation or third-party review depending on the system's risk category.

Answer

The system must be explainable using chain-of-thought reasoning

Answer

The system must achieve a minimum accuracy of 90% on validation data

Answer

The system must be trained on data that is representative of the target population

Question 20

A data science team wants to run a large-scale transformer training job with custom model architectures. They need the highest compute density for a multi-node job and want to minimize inter-node communication latency. Which Google Cloud infrastructure is BEST suited for this workload?

Accepted Answer

TPU v4 Pod slices with high-speed inter-chip interconnect. TPU v4 Pod slices provide the highest compute density for multi-node transformer training by using a custom 3D torus interconnect with 800 Gbps per chip bandwidth, which minimizes inter-node communication latency far below what standard networking can achieve. This architecture is specifically designed for large-scale model parallelism, making it ideal for custom transformer architectures that require frequent all-reduce and collective communication operations.

Answer

A single TPU v5e VM with multiple accelerators

Answer

A cluster of A100 GPU VMs connected via standard networking

Answer

Cloud Run jobs with GPU acceleration