Unit 3 roadmap — Techniques to improve gen AI model output

Under each topic, Google's own considerations are quoted from Google's Generative AI Leader exam guide, retrieved 2026-10-06; the numbered objectives beneath them are this hive's learning objectives, written over those considerations. Each topic closes with its own summary of what the exam actually tests, taken from its lecture.

Topic 1 — Recognize foundation model limitations and choose mitigations

GAIL-U3.T1 · 3 objectives · lecture deck of 10 slides

The exam guide's considerations, quoted. Identifying common limitations of foundation models (e.g., data dependency, the knowledge cutoff, bias, fairness, hallucinations, edge cases). Describing the Google Cloud-recommended practices to address limitations (e.g., grounding, retrieval-augmented generation [RAG], prompt engineering, fine-tuning, human in the loop [HITL]).

1. Identify data dependency, knowledge cutoff, bias, fairness, hallucination and edge-case limitations in a described output.
2. Match grounding, retrieval-augmented generation, prompt engineering, fine-tuning and human review to the limitation each addresses.
3. Decide when human-in-the-loop review is required before an output reaches a customer or a decision.

What this topic actually tests. Name it: outdated → knowledge cutoff; confidently wrong → hallucination; uneven across groups → bias and fairness; rare request misread → edge case. Treat it: grounding and RAG for facts, prompting for direction, fine-tuning for depth. Check it: rights at stake means a human reviews first.

Topic 2 — Monitor and evaluate gen AI in production

GAIL-U3.T2 · 3 objectives · lecture deck of 11 slides

The exam guide's considerations, quoted. Recognizing Google-recommended practices for continuous monitoring and evaluation of gen AI models (e.g., automatic model upgrades, key performance indicators, security patches and updates, versioning, performance tracking, drift monitoring, Agent Platform Feature Store).

1. Choose key performance indicators and evaluation methods for a gen AI application.
2. Plan for versioning, automatic model upgrades and security patches.
3. Explain drift monitoring and performance tracking, including the role of Agent Platform Feature Store.

What this topic actually tests. Measure: KPIs from business objectives, continuous output evaluation, endpoint health, human review. Change safely: versioned, regression-tested, released to a few first, with rollback — and patched. Notice: drift and quality shifts, caught by scheduled monitoring with threshold alerts; Feature Store keeps feature data consistent.

Topic 3 — Use prompt engineering techniques

GAIL-U3.T3 · 4 objectives · lecture deck of 12 slides

The exam guide's considerations, quoted. Defining prompt engineering and describing its significance in interacting with large language models (LLMs). Identifying prompting techniques and use cases (e.g., zero-shot, one-shot, few-shot, role prompting, prompt chaining). Identifying advanced prompting techniques and when to use them (e.g., chain-of-thought prompting, ReAct prompting).

1. Explain what prompt engineering is and why it matters when working with large language models.
2. Choose zero-shot, one-shot or few-shot prompting for a task.
3. Apply role prompting and prompt chaining to a business task.
4. Recognize when chain-of-thought or ReAct prompting fits a task.

What this topic actually tests. Is it tested? define the outcome and iterate. Does it need examples? zero-shot first; few-shot to fix a pattern. Is the job too big? a role sets behaviour; a chain splits the work. Does it need reasoning or action? chain-of-thought for steps; ReAct when an agent must act and observe.

Topic 4 — Ground outputs and control generation

GAIL-U3.T4 · 5 objectives · lecture deck of 14 slides

The exam guide's considerations, quoted. Describing the concept of grounding in LLMs and differentiating between grounding with first-party enterprise data, third-party data, and world data. Describing how retrieval-augmented generation (RAG) can affect the generated output from your gen AI models. Google Cloud grounding offerings: Identifying how sampling parameters and settings are used to control the behavior of gen AI models (e.g., token count, temperature, top-p [nucleus sampling], safety settings, and output length).

1. Distinguish grounding with first-party enterprise data, third-party data and world data.
2. Explain how retrieval-augmented generation changes a model's output.
3. Choose among prebuilt RAG with Agent Search, RAG APIs and Grounding with Google Search.
4. Predict the effect of temperature, top-p and token or output-length limits on generated output.
5. Explain what safety settings control and when to adjust them.

What this topic actually tests. Which data? first-party, third-party or world. Why RAG? retrieved facts become context — good retrieval in, grounded answer out. Which offering? prebuilt Agent Search, RAG APIs, or Google Search. Which settings? temperature and top-P for freedom, tokens for length, thresholds for safety.