Gemini Meetup Workshops

Welcome to the Gemini Meetup Workshops repository! This collection serves as a central hub for hands-on, educational, and experimental projects showcasing the power and versatility of Google's Gemini models and related AI technologies (such as CrewAI, PettingZoo, etc.).

Our goal is to provide practical, open-source examples that demonstrate cutting-edge AI applications, from intelligent agents playing games and generating content to embedding AI on edge devices. Whether you're a developer, AI enthusiast, or just curious, these workshops offer a fantastic way to dive deep into real-world Gemini implementations.

🎬 Interactive Demos

🎬 Cameo

AI-powered personalized video generator 📖 Documentation

♟️ Chess Companion

Real-time game analysis with voice interaction 📖 Documentation

🎬 TV Companion

AI film critic that controls your TV 📖 Documentation

🎮 Wind Waker Chat

Voice chat with controller support 📖 Documentation

🧱 Bricks

Fine-tuning Gemini for brick-building 📖 Documentation

🎨 Artifacts

AI-powered interactive content platform 📖 Documentation

🎯 Multi-Agent Games

Gemini plays games with PettingZoo 📖 Documentation

🥊 Smash Bot

LLM-powered fighting game agent 📖 Documentation

🍄 Mario Level Generator

AI creates Super Mario Bros levels 📖 Documentation (Coming Soon)

🖥️ Desktop Assistant

Gemini controls your desktop 📖 Documentation (Coming Soon)

🏈 Fantasy Football Draft Companion

Gemini guides your fantasy football draft 📖 Documentation

View All 18+ Workshops →

🎤 Talks & Presentations

Conference talks, interviews, and educational content exploring AI, LLMs, and the future of human-AI interaction:

📺 Interviews & Media

📚 Workshop Directory

Explore the diverse range of projects available in this repository. Each link takes you to a dedicated subdirectory with its own README.md containing detailed instructions and code.

adventure/ - Dungeon Adventure using CrewAI and Gemini: A text-based adventure game leveraging CrewAI and Gemini to create dynamic, agent-driven storytelling with a persistent Dungeon Master and multiple player agents.
artifacts/ - Unveiling Artifacts: AI-Powered Interactive Content Platform: Explores AI-driven content creation, demonstrating how LLMs can dynamically generate and refine interactive web experiences (HTML, CSS, JS) through an iterative, conversational workflow.
bricks/ - Bricks: Fine-tuning Gemini for Brick-Building: An end-to-end pipeline for fine-tuning Gemini 1.5 Pro to understand natural language descriptions and generate precise LDraw (MPD) brick-building instructions. (📊 slides)
cameo/ - Cameo: AI-Powered Personalized Video Generator: Creates personalized videos using your face and voice, combining MediaPipe face detection, Google Veo 3 video generation, and ElevenLabs voice cloning for fully customized AI-generated video content.
chess/ - Chess Companion: An AI-powered assistant providing real-time analysis, voice interaction, and TV control for live chess games, combining expert insights with historical context and engine analysis.
cost/ - LLM Cost Comparison: A hypothetical cost-comparison analysis of Gemini, GPT, and Llama 3 for a high-traffic app scenario, presented in a workshop slide format. (📊 slides)
doodles/ - Dynamic SVG Storytelling & Visualization: A powerful pipeline for converting static bitmap images into dynamic, interactive, and animatable SVGs for engaging visual storytelling and presentations.
factuality/ - Factuality: News Summarization & Redaction: A Perplexity-style clone built with CrewAI that summarizes, cites, and redacts news articles, focusing on factual accuracy and source verification. (📊 slides)
fantasy/ - Fantasy Football Draft Companion A Chrome extension and server combination that streams your screen and microphone input to Gemini for commentary over a Sleeper draft, as well as assistance with any user queries, player research, or auto-recommendations.
games/ - Gemini Plays Games: Multi-Agent Game Simulation: Demonstrates how to play multi-agent games with Gemini using PettingZoo and CrewAI, exploring AI interaction and strategy in simulated game environments. (📊 slides)
history/ - Interactive LLM History & Concepts Visualizer: Provides dynamic, animated visualizations to explain the historical development, growth patterns, and conceptual challenges (like "model collapse") of Large Language Models. (📊 slides)
kundali/ - Kundali Milan: AI Jyotish Guru: Implements a Jyotish guru as a CrewAI agent to perform Kundali Milan (Vedic astrology compatibility analysis) based on provided birth details.
podcast/ - AI Podcast Generator: An innovative, AI-driven pipeline for automatically generating engaging podcast-style video content, from dynamic scripting and voice-overs to music integration and final video assembly.
prisoner/ - LLM Agents in Game Theory: Prisoner's Dilemma Simulator: A fascinating exploration into AI strategy, simulating the classic Prisoner's Dilemma with two distinct LLM agents strategizing and interacting in an evolving, turn-based game.
recap/ - Workshop Recap: A summary of key learnings and advancements made since the inception of the Gemini workshops, presented as a slide deck. (📊 slides)
roast/ - AI Roast Battle Arena: Uncensored Wit with LLMs: Showcases a dynamic, multi-agent AI system designed to host uncensored roast battles between different LLMs, with an independent third LLM acting as an impartial judge.
scripts/ - LLM-Powered Content Generation & Developer Experience: A robust framework for generating various content types, including presentation slides and factual summaries, powered by LLMs, emphasizing efficient developer workflows.
startupcon/ - StartupCon Talk: Workshop presentation on building with Gemini and AI agents. (📊 slides)
smash/ - Smash Bot: An LLM-Powered Fighting Game Agent: A cutting-edge AI agent capable of playing a fighting game (like Super Smash Bros.) by observing the game state and controlling a virtual gamepad, powered by a multimodal LLM. (📊 slides)
tv/ - TV Companion: Your AI Film Critic and Controller: An advanced LLM-powered assistant that "watches" movies and shows in real-time, offering insightful commentary, deep-diving into film trivia, and controlling your TV through natural voice commands.
utils/ - LLM Demonstration Workshop Project: A practical demonstration of integrating LLMs into applications, showcasing advanced capabilities through automated content generation, such as real-time news summaries and LLM cost comparisons.
waker/ - Wind Waker Voice Chat with Controller Support: A real-time AI gaming companion for The Legend of Zelda: Wind Waker, combining native audio voice chat, visual game understanding, direct controller actuation, and episodic memory. (📊 slides)
wearable/ - Wearable Gemini: Demonstrates how to embed Gemini on a Raspberry Pi Zero W with an Adafruit Voice Bonnet, utilizing Google's speech-to-text and text-to-speech for an on-device AI assistant.

🚀 Getting Started

To explore and run these workshops:

Clone the Repository:

git clone https://github.com/google-gemini/workshops.git
cd workshops

Navigate to a Project: Choose a workshop from the directory above and navigate into its specific folder (e.g., cd adventure/).
Read the Project's README: Each subproject has its own README.md with detailed instructions on prerequisites, installation, API key setup, and how to run the specific demonstration.
Install Dependencies: Most projects will require Python and dependency management tools like pip or poetry. Follow the README.md for the exact steps.
Set Up API Keys: Many workshops utilize the Gemini API and potentially other Google Cloud services (Speech-to-Text, Text-to-Speech). Ensure you have the necessary API keys and credentials configured as described in the individual project's README.md.

🎯 Categories

To help you find projects of interest, we've grouped them into thematic categories. Note that some projects may naturally fit into multiple categories due to their multifaceted nature.

AI/ML Applications & Agents

Projects focusing on practical applications of LLMs, multi-agent systems, and real-world problem-solving.

adventure/ - Dungeon Adventure using CrewAI and Gemini
artifacts/ - Unveiling Artifacts: AI-Powered Interactive Content Platform
bricks/ - Bricks: Fine-tuning Gemini for Brick-Building
factuality/ - Factuality: News Summarization & Redaction
kundali/ - Kundali Milan: AI Jyotish Guru
roast/ - AI Roast Battle Arena: Uncensored Wit with LLMs
tv/ - TV Companion: Your AI Film Critic and Controller
utils/ - LLM Demonstration Workshop Project

Games & Interactive Entertainment

Projects where AI interacts within game environments, provides interactive entertainment, or simulates game theory scenarios.

chess/ - Chess Companion
fantasy/ - Fantasy Football Draft Companion
games/ - Gemini Plays Games: Multi-Agent Game Simulation
prisoner/ - LLM Agents in Game Theory: Prisoner's Dilemma Simulator
smash/ - Smash Bot: An LLM-Powered Fighting Game Agent
waker/ - Wind Waker Voice Chat with Controller Support

Content & Media Generation

Projects focused on generating various forms of content, including text, visual, audio, and video, using AI.

artifacts/ - Unveiling Artifacts: AI-Powered Interactive Content Platform
bricks/ - Bricks: Fine-tuning Gemini for Brick-Building
cameo/ - Cameo: AI-Powered Personalized Video Generator
doodles/ - Dynamic SVG Storytelling & Visualization
factuality/ - Factuality: News Summarization & Redaction
podcast/ - AI Podcast Generator
scripts/ - LLM-Powered Content Generation & Developer Experience
utils/ - LLM Demonstration Workshop Project

Tools, Utilities & Core Concepts

Projects that build development tools, illustrate fundamental AI concepts, provide analytical insights, or enhance developer experience.

cost/ - LLM Cost Comparison
history/ - Interactive LLM History & Concepts Visualizer
recap/ - Workshop Recap
scripts/ - LLM-Powered Content Generation & Developer Experience
utils/ - LLM Demonstration Workshop Project

Hardware & Edge AI

Projects demonstrating the integration of AI capabilities on physical hardware or embedded systems.

wearable/ - Wearable Gemini

We encourage you to explore, experiment, and contribute! Happy learning!