What's changed: Created Cloud Digital Leader Chapter 3 (Domain 3 "AI": AI/ML fundamentals = AI/ML/generative AI relationship, supervised/unsupervised, training data, responsible AI; Google Cloud AI/ML solutions = Vertex AI/Gemini/Vision/Speech-to-Text/Natural Language/Translation/BigQuery ML/Document AI/Conversational Agents).
3.2Google Cloud AI/ML solutions
Understand the roles and selection of Google Cloud AI/ML solutions: Vertex AI for building and using AI/ML, Gemini at the core of generative AI, ready-to-use pretrained APIs (Vision, Speech-to-Text, Natural Language, Translation), BigQuery ML for SQL-based machine learning, Document AI for document processing, and Conversational Agents for customer service.
Google Cloud lets you use AI at various levels of maturity, from "build your own model" to "just call a finished API." You choose based on how much AI expertise your organization has.
3.2.1Vertex AI and Gemini
Vertex AI is a unified AI platform that handles building, training, deploying, and operating ML models end to end—from developing custom models to using Google’s foundation models. Its core generative AI model is Gemini, which generates text, summarizes, assists with code, and understands images. Gemini is also embedded across Google Cloud services and Google Workspace to assist work. Teams with sufficient expertise build custom models on Vertex AI; otherwise, using pretrained APIs or Gemini directly is the fast path.
3.2.2Ready-to-use AI solutions
| Solution | What it does | Example use case |
|---|---|---|
| Vision API | Classify images, detect objects/text | Auto-tag product images |
| Speech-to-Text | Convert speech to text | Transcribe call center audio |
| Natural Language API | Analyze sentiment/entities in text | Sentiment analysis of reviews |
| Translation API | Translate across languages | Localize a website |
| Document AI | Extract data from documents | Process invoices/contracts |
| Conversational Agents | Conversational agents/chatbots | Automate customer support |
For data analytics teams, BigQuery ML lets you create ML models and run predictions with just SQL on data in BigQuery. The benefit is doing demand forecasting and more in familiar SQL without moving data. From the Cloud Digital Leader perspective, these "ready-to-use solutions" matter because they deliver business value quickly even without AI specialists.
For building custom models more deeply, the open-source ML framework TensorFlow is widely used, and Cloud TPU (Tensor Processing Unit) is Google's purpose-built hardware that accelerates its training and inference. As a leader you need not memorize details, but grasp the big picture that "large-scale AI is powered by frameworks like TensorFlow and specialized hardware like the TPU."
Common: goal → solution. E.g., "build/operate custom models" = Vertex AI; "generative AI like text generation/summarization" = Gemini; "image recognition" = Vision API; "speech → text" = Speech-to-Text; "translation" = Translation API; "sentiment analysis of text" = Natural Language API; "extract data from invoices" = Document AI; "chatbot for customer service" = Conversational Agents; "ML with just SQL" = BigQuery ML.
Watch the mix-ups: (1) Vertex AI (platform to build/operate) vs pretrained APIs (just call) differ in maturity. (2) Gemini = generative AI model; Vertex AI = the platform to use them. (3) Task-specific APIs (Vision/Speech/Translation/NL/Document AI) have fixed roles. (4) BigQuery ML does ML in SQL without moving data.
3.2.3Section summary
- Vertex AI = unified platform to build/operate models; Gemini = core generative AI model
- Pretrained APIs = Vision/Speech-to-Text/Natural Language/Translation; documents = Document AI; conversation = Conversational Agents
- BigQuery ML = machine learning in SQL without moving data; fast value even without specialists
Sign in to track progress — Log in.
Quick check
(just a quick review)Q1. Which Google Cloud unified AI platform builds, trains, deploys, and operates ML models end to end?
Q2. Which Google Cloud API converts audio into text?
Q3. Which API best classifies images and detects objects or text?
Q4. Which lets you create ML models and predict using only SQL on data in BigQuery, without moving it?
Q5. Which Google Cloud solution best extracts information from documents like invoices and contracts?
Q6. Which is Google Cloud’s core generative AI model for text generation, summarization, and code assistance?