Google and Kaggle Rerun Their Gen AI Intensive as a Five-Day Live Course

What the 2025 course actually adds

The core announcement is a repeat with modifications. Google and Kaggle ran the Gen AI Intensive in 2024 and report more than 140,000 participants. The 2025 edition runs on a fixed schedule, March 31 through April 4, and Google says it brings updated content, new speakers, and one structural addition: a Kaggle capstone project at the end.

The stated curriculum covers foundational topics — prompt engineering and embeddings are the two named explicitly — delivered through daily livestreams where Google experts answer questions in real time. That five-day, time-boxed format matters. It is a cohort model, not an on-demand library, which is why the specific dates carry weight in the announcement.

The materials themselves are a mix: whitepapers written by Google experts, hands-on code labs using Gemini and other Google services, a dedicated Discord for community and expert contact, and AI-generated podcasts produced with NotebookLM. That last detail is notable — Google is using one of its own gen AI products as a content-production tool inside a course about gen AI.

The capstone changes what the course measures

In 2024 the course was primarily instructional. Adding a Kaggle capstone project shifts the endpoint from attendance to demonstrated application. Participants apply the learned skills to a practical gen AI challenge, and the announcement pairs this with a certificate and the chance to compete for prizes.

Kaggle is the operative venue here. It is already a platform where work is submitted, scored, and ranked. Routing the capstone through Kaggle means the course produces artifacts and leaderboard-style outcomes rather than just completion checkmarks. For anyone evaluating a candidate, a Kaggle capstone is more verifiable than a passive certificate.

Free, but built around Gemini

The announcement is explicit that the course remains free for everyone. That is the headline incentive, but the code labs are the mechanism that matters. They give hands-on experience with Gemini and other Google services, which means 140,000-plus practitioners spend a week building against Google's API surface at no cost to them.

Coursework will include AI-generated podcasts using NotebookLM, whitepapers by Google experts and practical code labs for hands-on experience with Gemini and other services.Montana Labs

There is nothing hidden in this — a free course taught by the vendor uses the vendor's tools. But it is worth naming plainly: the free tuition is offset by developer familiarity with Gemini, which becomes the default reference point for participants who later choose models and platforms at work.

The implication: a recurring funnel from newcomer to Gemini-fluent builder

By repeating the course a second year and adding a capstone, Google is treating the Gen AI Intensive less as a one-off event and more as a standing pipeline. The structure moves someone from foundational concepts through hands-on Gemini code labs to a submitted Kaggle project in five days, with a certificate and a Discord community at the end.

For applied teams, the practical read is simple. This is a cheap, structured way to bring engineers up to a working baseline on prompt engineering, embeddings, and Gemini tooling within a fixed week — and to get a Kaggle artifact that shows they actually built something. The tradeoff is that the fluency it produces is Google-flavored, so the skills transfer best inside a Gemini-centric stack.