Agentic Workflows with Google ADK

Übersicht

In "Agentic Workflows with ADK," learners will explore the principles of creating intelligent, automated workflows using the Google Agent Development Kit and Vertex AI Gemini. The course begins by defining agentic workflows and progresses into hands-on lessons for modeling and implementing various workflow patterns, including prompt chaining, routing, and parallelization.

Participants will also tackle more complex patterns like evaluator-optimizer and orchestrator-worker workflows. The course culminates in a project where students will create an AI Research Assistant, applying their skills to develop a sophisticated agent-based system.

Ideal for those interested in advancing their knowledge of AI and workflow automation.

Lehrplan

• Welcome to Agentic AI Workflows with Google Agent Development Kit
Explore agentic AI workflows and get started with Google's Agent Development Kit by learning prerequisites, setup, and essential concepts.
• Understanding Agentic Workflows
Explores what defines a modern AI agent, its core components (Persona, Knowledge, Tools, Interaction), and the different types of agents based on their LLM interaction model.
• Implementing Agentic Workflow Analysis and Definition with ADK
Learn to implement agentic workflows using ADK by decomposing tasks and coordinating analyzer, coordinator, and validator agents for scalable, validated IT processes.
• Agentic Workflow Modeling
Design and visualize agentic workflows. Learn common agent types as building blocks for creating visual workflow diagrams.
• Implementing Agentic Workflow Modeling with ADK
Learn to design and implement agentic workflows in ADK using sequential, conditional, and parallel patterns, plus orchestration, error handling, and metrics for optimization.
• Agentic Workflow Implementation
Covers the practical aspects of translating agentic workflow models into Python code. Students learn to structure agent logic, define agent classes, and orchestrate their interactions.
• Implementing Agentic Workflows with ADK and Vertex AI Gemini
Explore agentic workflow modeling with ADK and Vertex AI Gemini: manage states, visualize flows, handle errors, and analyze execution for robust multi-agent automation.
• Agentic Workflow Patterns: Prompt Chaining Workflow
Introduces the Prompt Chaining pattern for breaking down complex tasks into a sequence of smaller, dependent steps. It covers strategies for task decomposition, validation, and context management.
• Implementing Prompt Chaining Agentic Workflows with ADK and Vertex AI Gemini
Learn to implement multi-step agentic workflows with ADK, integrate Vertex AI Gemini LLM, use sequential and parallel patterns, and test agent performance.
• Agentic Workflow Patterns: Routing
Teaches the Routing pattern, which involves classifying incoming tasks and directing them to the most appropriate specialized agent or processing path.
• Implementing Routing Agentic Workflows with ADK and Vertex AI Gemini
Learn to build iterative agentic workflows using ADK and Vertex AI Gemini, implementing loop agents for automated refinement with generator and critic agents, escalation logic, and safe termination.
• Agentic Workflow Patterns: Parallelization
Introduces the Parallelization pattern for executing multiple agent tasks concurrently. It covers strategies for task decomposition (sharding, aspect-based) and result aggregation.
• Implementing Parallelization Agentic Workflows with ADK and Vertex AI Gemini
Learn to implement parallel agentic workflows using ADK and Vertex AI Gemini with fan-out/fan-in patterns for efficient concurrent task execution and automated result aggregation.
• Agentic Workflow Patterns: Evaluator-Optimizer Workflow
Focuses on the Evaluator-Optimizer pattern, an iterative process of generation, critique, and refinement to improve output quality. It emphasizes clear evaluation criteria and actionable feedback.
• Implementing Evaluator-Optimizer Workflows with ADK and Vertex AI Gemini
Learn to implement intelligent routing in ADK with Vertex AI Gemini, classifying content, selecting specialist agents, and building robust, rule-based workflows for multimodal data.
• Agentic Workflow Patterns Orchestrator-Workers Workflow
Introduces the advanced Orchestrator-Workers pattern, where a central agent dynamically plans, delegates, and synthesizes the work of multiple specialized worker agents.
• Implementing Orchestrator-Worker Workflows with ADK and Vertex AI Gemini
Explore how to design and implement orchestrator-worker workflows using ADK and Vertex AI Gemini for scalable AI task management.
• Project: AI Research Assistant with Google Agent Development Kit and Vertex AI Gemini
Build a complete AI Research Assistant that analyzes research queries, routes to specialized agents, executes parallel workflows, and generates comprehensive reports.

---

Unterrichtet von

Peter Kowalchuk and Noble Ackerson

---

Fachgebiete

Artificial Intelligence