Hands-on: Building a Semantic Search System with Gemini Embedding 2
Hands-on with Gemini Embedding 2: run embeddings in Colab and build a small local semantic search system to learn how RAG works.
- When
- Thu, May 14, 2026 · 19:00–21:30 JST
- Where
- Google 渋谷オフィス · In person
- Region
- Kanto (Tokyo)
- Organizer
- Google Developer Group Tokyo
- Language
- JA
- Source
- Connpass
Summary
A Build with AI hands-on hosted by Google Developer Groups Tokyo. Using Gemini Embedding 2 (released in March 2026), participants run small text and image embedding examples to build intuition for how embeddings and RAG work.
The second half walks through a local sample of a lightweight semantic search system so attendees can see a complete retrieval pipeline end to end. The basic track only requires a personal Google account and Colab access, while the advanced track expects git, uv, and Python 3.12+.
The speaker is tsho (Sho Tanaka), Lead Developer Advocate at Snowflake and a Google Developer Expert in AI/ML. The session is held at the Google Shibuya office, free to attend, with an optional networking slot at the end.
About the community
A Tokyo-based developer community focused on Google technologies. It runs study sessions, hands-on workshops, and DevFest events on weekday evenings and weekends, with a recurring Build with AI generative-AI workshop series. This session is co-hosted with the student sister community GDGoC UTokyo.
#ai#gemini#embeddings#rag#semantic-search#hands-on#google#generative-ai#machine-learning