The Agentic AI Digest (14 Nov) | ADK Quickstart, Gemini's File Search (RAG) & Agent Builder Updates

This week: We explore the new Agent Development Kit (ADK) for Go, and highlight Gemini's new File Search tool and major updates to Vertex AI Agent Builder.

Hi everyone,

Welcome to your weekly briefing from the Agentic AI Roundtable. Our goal is to cut through the noise and deliver the most relevant signals, patterns, and community wins to help you build more effectively.

Let’s dive in.

🛠️ Community Commits: Building in the Open

This week we explore the Agent Development Kit (ADK) QuickStart using the recently released GoLang SDK (see the “On the Radar” section below for details). ADK is a core component of Google’s Agent Builder stack for developing and deploying powerful agent applications. Exciting news that GoLang developers now get to also unlock these capabilities.

The QuickStart run-through sets up a simple “hello world” agent application powered by Gemini 2.5 Flash model and uses one of ADK’s built-in tools (Google Search) to answer user queries (much like a human would do if they needed to Google search some unknown information). To interact with the agent, we use ADK’s out-of-the-box web UI for a user-friendly interface and visibility into some of the agent’s internal operations while executing. Keep a lookout for more advanced use-cases with ADK in the coming weeks.

ADK GoLang QuickStart

Useful links

• ADK project

📡 On the Radar: What’s Moving the Needle

A curated look at the articles, papers, resources and updates that are worth your time this week.

• Google has announced a major expansion of its agent-building capabilities, focusing on the full lifecycle of production agents. The Gemini API now features a File Search tool, a fully-managed RAG system that handles file storage, chunking, and embeddings to ground responses in your data. Concurrently, Vertex AI Agent Builder received new features to “build, scale, and govern,” including better context control in the ADK, a “self-healing” tool plugin, new observability and evaluation dashboards in the Agent Engine, and native agent identities for secure IAM-based governance.

  • Introducing the File Search Tool in Gemini API

  • More ways to build and scale AI agents with Vertex AI Agent Builder

• The Agent Development Kit (ADK) and Agent-to-Agent (A2A) protocol are now available for Go. This allows developers to leverage Go’s low-latency startup times, high performance, concurrency, and strong typing to build agents.

  • Announcing the Agent Development Kit for Go

  • ADK GoLang Repository

  • A2A GoLang Repository

• For builders looking to solidify their foundational knowledge, Google has published two essential whitepapers on agentic AI. The first, an “Introduction to Agents,” frames the paradigm shift from passive models to autonomous, problem-solving applications. The second, on “Context Engineering, Sessions, and Memory,” provides a deep dive into the critical techniques for building stateful agents that can remember, learn, and personalize interactions.

  • Kaggle Whitepaper: Introduction to Agents

  • Kaggle Whitepaper: Context Engineering, Sessions, and Memory

🤝 Want to Get Involved in the Community?

This roundtable is driven by its members. To join the conversation, share your work, or ask a question, you have two great options:

1. Join our private Google Chat space for real-time discussions and to participate in the weekly Open Thread. [Link to Chat Space]

2. Send a message to our community Google Group at roundtable-community@agentic-ai.build.

We look forward to hearing from you.

The Agentic AI Roundtable Core Team