Google and Kaggle announced a new free intensive course on AI Agents, Vibe Coding and autonomous workflows. The program runs June 15–19, 2026 and follows on from the previous GenAI Intensive, which crossed 1.5 million participants globally.
Dates, cost and where to register
- Dates: June 15–19, 2026
- Cost: free
- Official course page (Kaggle): 5-Day AI Agents Intensive — VibeCoding Course with Google
- Official announcement: Google blog post
What you need to participate
- an active Google account (Gemini, Google AI Studio, etc.)
- a Kaggle account — create one here
- basic familiarity with Python and LLM APIs
- an environment to run notebooks (Kaggle Notebooks is enough)
How to register — step by step
- Open the official Kaggle page.
- Click Join Competition / Register (the label varies with the current Kaggle UI).
- Accept the participation terms and rules.
- Join the course Discord — networking, support and livestreams with Google engineers.
Program structure
- Daily assignments: Kaggle Jupyter Notebooks (
.ipynb) with theory, exercises and hands-on labs every day — everything runs in the Kaggle cloud, no local install required - Daily livestreams: demos, architecture deep dives and live Q&A with Google engineers
- Capstone project: a complete final agent-based project
What is Vibe Coding?
The term Vibe Coding describes a paradigm in which natural language becomes the primary interface for building software.
Instead of manually implementing every detail, engineers describe:
- objectives
- expected behavior
- business rules
- integrations
- workflows
- architectural constraints
And AI models generate a significant portion of the technical implementation.
In practice, the developer’s role shifts from:
- manual implementer
- boilerplate writer
- operational executor
toward:
- intelligent systems architect
- agent orchestrator
- autonomous workflow supervisor
- context designer
The rise of agentic systems
Google’s announcement makes the direction explicit: the industry has moved far beyond simple “code autocomplete.”
The focus is now on:
- AI Agents
- tool orchestration
- reasoning workflows
- persistent memory
- autonomous execution
- agent collaboration
- multi-step planning
Tools such as:
- Cursor
- Claude Code
- OpenAI Codex
- Devin
- Windsurf
already show this evolution by letting developers operate multiple agents in parallel.
What the course covers
The program covers core topics from the next generation of software engineering.
Building AI Agents
Agents capable of:
- reasoning
- executing tasks
- making decisions
- using tools
- consuming APIs
- operating complex workflows
Multi-Agent Systems
The course also explores architectures composed of multiple specialized agents collaborating with one another.
Common patterns:
- planner agents
- supervisor agents
- executor agents
- reviewer agents
- critic agents
These architectures are increasingly common in modern frameworks such as:
- LangGraph
- CrewAI
- Semantic Kernel
- OpenAI Agents SDK
Tool Calling and Autonomous Workflows
Another major focus:
- API integrations
- persistent memory
- agentic workflows
- inter-agent communication
- context-driven execution
- orchestration pipelines
Concepts becoming foundational in modern agent-based platforms.
Hands-on project
By the end of the course, participants build a project based on agents wired into real tools and autonomous workflows.
The goal is to expose students to scenarios close to modern production-grade AI engineering.
The structural shift in software engineering
The announcement reinforces a meaningful structural transformation across the technology industry.
For decades, software engineering largely meant:
- writing code
- manually implementing logic
- building integrations line by line
Now the focus is shifting toward:
- context definition
- agent-oriented architecture
- intelligent workflow design
- autonomous execution supervision
- governance of AI-driven systems
Code remains important, but increasingly becomes an implementation layer that is progressively automated.
Opportunity for senior engineers
Despite the massive productivity gains enabled by LLMs, the rise of Vibe Coding introduces real challenges:
- reliability
- observability
- security
- debugging
- governance
- operational cost
- reasoning validation
This tends to raise the value of professionals with strong backgrounds in:
- distributed architecture
- resilient systems
- platform engineering
- DevOps
- Kubernetes
- cloud-native systems
- production engineering
In this landscape, technical seniority is no longer just “writing code fast.” It is increasingly tied to the ability to:
- design intelligent systems
- coordinate agents
- validate behavior
- control autonomous execution in production
Conclusion
The new Google and Kaggle course shows that AI Agents and Vibe Coding are no longer just emerging trends — they already represent a concrete transformation in modern software engineering.
The industry is rapidly moving toward a model where:
- agents execute tasks
- workflows self-orchestrate
- models make contextual decisions
- developers act as intelligent systems architects
More than a tooling evolution, this looks like a paradigm shift comparable to the rise of cloud computing, containers and DevOps.
If you work with software development, architecture, automation or applied AI, this is one of the best free entry points right now to understand how AI Agents and Vibe Coding are redefining modern software engineering.