Overview of Generative AI on Google Cloud | Austin .

Members-Only

Recent Talks & Demos are for members only

Exclusive feed

You must be an AI Tinkerers active member to view these talks and demos.

November 10, 2025 · Austin

Google Cloud GenAI and Gemini

This talk covers Generative AI on Google Cloud, detailing ADK, Agent Engine, Evaluations, GenMedia, and the new Gemini Computer Use model.

Overview
Tech stack
  • ADK
    ADK (Agent Development Kit) is Google's open-source, code-first framework: it builds, orchestrates, and deploys sophisticated, production-ready AI agents across Python, Java, and Go.
    The Agent Development Kit (ADK) is an open-source, code-first framework from Google, designed to simplify the full-stack development of intelligent, multi-agent systems. It provides the core architecture (orchestration, state management) necessary to move beyond single-purpose models: developers use it to compose specialized agents in a hierarchy for complex, multi-step workflows. While optimized for the Gemini model and Google Cloud's Vertex AI Agent Engine deployment, ADK maintains model-agnosticism and supports multiple languages (Python, Java, Go), ensuring scalability and flexibility for enterprise-grade AI applications.
  • Agent Engine
    Agent Engine: The managed service on Vertex AI for deploying, scaling, and governing enterprise-grade AI agents in production.
    This is your production launchpad for autonomous agents: Agent Engine (part of Vertex AI Agent Builder) handles the heavy lifting. It's a set of services that manages infrastructure, scaling, security, and monitoring, freeing developers to focus on core agent capabilities. Agents deployed here leverage features like short-term and long-term Memory Bank for context-aware, human-like interactions. You get full control over the agent lifecycle, from integrated evaluation and Gemini model support to seamless deployment of agents built with frameworks like LangChain or CrewAI.
  • Evaluations
    DeepEval is the open-source LLM evaluation framework: it functions as a Pytest-like unit testing tool for validating large language model outputs with programmatic rigor.
    Evaluations, specifically via the DeepEval framework, provide the necessary structure for systematic LLM testing. This open-source tool integrates directly into your CI/CD pipeline, acting like a specialized Pytest for AI applications. It leverages over 50 research-backed metrics—including G-Eval, RAGAS, and Hallucination checks—to score model performance on specific criteria. Developers define test cases, run the evaluation, and receive concrete metrics to prevent regressions, ensuring model reliability before deployment.
  • GenMedia
    GenMedia Partners is an integrated sales and marketing powerhouse, providing end-to-end solutions for broadcasters and advertisers across multiple platforms (radio, digital, out-of-home, events)
    GenMedia Partners is an integrated sales and marketing powerhouse, providing end-to-end solutions for broadcasters and advertisers across multiple platforms (radio, digital, out-of-home, events). We leverage advanced tools for precision media buying and campaign execution. Our reach is significant: we provide 98% U.S. national spot coverage and deliver to over 82% of the U.S. Hispanic population via HRN Media Network. We drive efficiency with proprietary systems like the Rapid Rate System and Avail Alert System, ensuring our clients — from major agencies to broadcasters — maximize ROI and revenue growth.
  • Gemini
    Google's natively multimodal AI model: understands and operates across text, code, audio, image, and video.
    Gemini is Google's most capable and general AI model, engineered from the ground up to be natively multimodal: it seamlessly understands and combines information across text, code, audio, image, and video inputs. The technology is optimized for flexibility, running efficiently on everything from data centers to mobile devices. It is deployed in three key sizes: Ultra (for highly complex tasks), Pro (for broad scaling), and Nano (for efficient on-device tasks). Developers access this power via the Gemini API to build next-generation applications.

Related projects