Google CAN'T STOP COOKING: FREE AI AGENTS SDK JUST DROPPED

preview_player
Показать описание
Someone NEEDS TO GET GOOGLE OUT THE KITCHEN IMMEDIATELY - they're changing the AI game, first with Gemini 2.5 pro, then firebase studio, and now This FREE Agent Development Kit (ADK)

Google just released their Agent Development Kit (ADK) and it's a GAME CHANGER! In this video, I break down everything you need to know about this powerful new tool that allows you to create AI agents with dynamic routing and decision-making capabilities. The ADK works with all popular LLMs (including Anthropic's Claude) and features built-in Google Search, code execution, and MCP tools integration. I'll show you how to get started, explore the key features, and explain why this might be the most exciting AI development tool released yet. With just a few lines of code, you can build agents that autonomously execute workflows, make decisions, and connect to third-party services like Postgres, Salesforce, and more. Google is absolutely COOKING right now with Gemini 2.5 Pro, Firebase Studio, and now the ADK!

Join the Skool community today:

#GoogleADK #AIAgents #AgentDevelopmentKit #Gemini25Pro #AItools #GoogleAI #DynamicRouting #MCPtools #AIOrchestration #aiworkflows

Chapters

0:00 - 0:20 Introduction to Google's Agent Development Kit release
0:20 - 0:46 Overview of ADK compatibility with various LLMs
0:46 - 1:24 Getting started with pip install and initial setup
1:24 - 2:09 Flexible orchestration: sequential vs. dynamic routing
2:09 - 2:40 Multi-agent architecture and coordination features
2:40 - 3:05 Rich tool ecosystem and MCP integration
3:05 - 3:32 Deployment options and containerization
3:32 - 4:05 Setting up your first ADK project
4:05 - 4:52 How agents make decisions with dynamic routing
4:52 - 5:29 Web UI, terminal, and API server interaction options
5:29 - 6:00 Streaming responses and interface capabilities
6:00 - 6:38 Built-in Google Search tool functionality
6:38 - 7:24 Code execution capabilities for data processing
7:24 - 8:03 Vertex AI search for private data stores
8:03 - 8:47 Third-party tool integrations (Langchain and more)
8:47 - 9:35 Google Cloud tools and enterprise connectors
9:35 - 10:02 MCP tools and Google Cloud integration
10:02 - 10:54 Using Roo/Cline to read documentation and build with ADK
10:54 - 11:35 Authentication options for various APIs
11:35 - 12:24 How LLM agents work as the "thinking part" of applications
12:24 - 13:13 Model compatibility with Claude and other providers
13:13 - 13:48 OLLAMA integration for free agent deployment
13:48 - 14:12 Final thoughts and future development plans
Рекомендации по теме
Комментарии
Автор

I am a developer, , but I don't feel like I need to develop any more. And I feel that your channel is probably the most informative no BS channel out there these days.

JasonMichael
Автор

Join the Skool community today:


Incomestreamsurfers
Автор

Another great vid thanks. Organising groups of agents is the way to go — add a self-learning neural net and some reflexive logic, it may discover more then its builder knows about whatever task it’s doing.

davidmiranski
Автор

Could you please post the videos in 30fps? I binge watch your content on Apple TV and a YouTube app on the smart TV and they don't support 4K/HD at 60fps together with higher speed playback. 60fps can only play with 1x speed. Thank you!

dale-ys
Автор

Heck, I was just wrapping up for the night...

🤣

Thanks Hamish

AiWatchTower
Автор

Is this a tutorial where you show us how to build stuff ... Or this is a video where we see you learn and comment on your own learning process ?

Rami_Zaki-kb
Автор

Can jt build Apps and Gemini 3.0 coming

holdthetruthhostage
Автор

Can you guess when could we get gemini 3.0?

ashishpraveen
Автор

So it's like crawl4ai but for... Literally everything, not just scraping

atetraxx
join shbcf.ru