Vertex AI Agent Builder: What It Is, with a Slack Search Example

Andreas Nigg
Andreas Nigg · Co-Founder, Pondhouse Data
Published Updated 10 min read
Vertex AI Agent Builder: What It Is, with a Slack Search Example
On this page

NOTE: This article is an introduction to the Google Vertex AI Agent Builder. We are not affiliated with Google. This article is also not to be understood as a review, but informational material as part of our first look experience.

Vertex AI Agent Builder is Google Cloud's platform for building AI agents and search applications that are grounded in your own data, mostly through a no-code console. In this post, we use it to build a RAG search engine on top of Slack messages: we connect Slack as a data store, create a Search app with Gemini-powered summaries, and embed it into a website as a widget.

The whole setup needs no code. With all LLM features enabled, it cost $8 per 1000 queries at the time of writing. And while it was easy to set up, the result is, in the end, just search.

Google launched Vertex AI Agent Builder in 2024 as part of the Google Cloud Vertex AI suite. It integrates with existing products like Vertex AI Search and Conversation. It's Google's first major, serious entry into RAG, which is surprising: a good chunk of RAG (Retrieval Augmented Generation) is about searching and finding information, and that's something Google is famously good at. So I was quite excited to test the Agent Builder.

Update (November 2026): Google has renamed the product used in this post several times. The original Vertex AI Agent Builder became AI Applications in April 2025, then Vertex AI Search, and since April 2026 Agent Search, although the console still says "Vertex AI Search" and "AI Applications". The name "Agent Builder" now refers to a different set of agent tools, and "LangChain on Vertex AI" became Vertex AI Agent Engine. Current pricing is $1.50 per 1000 queries (Standard) or $4 (Enterprise, now including basic generative answers), plus $4 for Advanced Generative Answers, and Slack data stores are now documented under Gemini Enterprise.

What is the Google Vertex AI Agent Builder?

Vertex AI Agent Builder is a platform that helps developers build and deploy AI agents, regardless of their level of expertise. It offers a no-code console for creating agents using natural language, as well as more advanced, open-source frameworks for experienced developers.

One of its main features is grounding: it keeps AI outputs factually accurate and relevant by grounding them in enterprise data. It does this through systems like Vertex AI Search and Retrieval Augmented Generation (RAG) APIs, as well as custom solutions for more complex needs.

Security and compliance matter too. Agent Builder adheres to industry standards like HIPAA and the ISO 27000 series. This alone is quite powerful, as it enables enterprise-grade applications that are otherwise underserved.

What's quite unique is that the platform guides users through each step of agent creation, from defining parameters to training the model, which makes development more accessible.

According to Google, AI agents can streamline business processes by automating routine tasks, freeing up resources and improving productivity. That's obviously a marketing sentence, but it is grounded in reality: AI agents can help with these tasks.

One of the main value propositions of the Vertex AI Agent Builder is its low-code user interface. Because it makes AI tools more accessible, Vertex AI Agent Builder could be a good way for businesses to get started with AI, even without much AI experience.

Finally, the integration with Google Cloud infrastructure means you have the full Google Cloud Platform at your fingertips for developing, deploying and managing your AI agents. If you want to use other models from the same platform, see our practical example of the Vertex AI Model Garden.

Summary: features of the Vertex AI Agent Builder

  • No-code AI agent creation: develop conversational AI agents without writing code, using natural language inputs.

  • Data grounding: connect AI agents to trusted data sources like Google Search and enterprise data using APIs, for accurate and relevant responses. Google provides connectors to some external tools for a full RAG experience.

  • Testing and deployment: tools for performance evaluation and tuning, with scalable deployment on enterprise-ready infrastructure.

  • Security and compliance: AI operations adhere to HIPAA and ISO standards, with data privacy management and responsible AI model usage.

  • Advanced AI capabilities: document parsing, ranking, grounded generation and fact-checking APIs for custom solutions.

  • Scalable vector search: applications that handle billions of vectors, combining vector and keyword search for relevant responses.

  • LangChain integration: use the open-source Python framework to create and deploy high-performance AI agents on Vertex AI. This part is especially interesting, as it lets you use some parts of the Google ecosystem and extend them with building blocks from other AI platforms. For example, you could use the Agent Builder connectors and vector search together with an OpenAI model.

How to connect the Vertex AI Agent Builder to your Slack messages

Now that we know what the Vertex AI Agent Builder can do, let's put it to the test.

We are going to build a RAG system that answers questions from our Slack messages. Why Slack? Slack is one of the tools where a company stores enormous amounts of knowledge. If we can make this knowledge accessible with AI, we create a lot of value.

Before we start, we need to set up access control for our data source. This topic is admittedly quite cumbersome. You can find more information in the Agent Builder data source access control documentation. To just get it done, navigate to the Agent Builder authentication settings, click the "edit" button and select "Google Identity" as the identity provider for your data sources.

  1. Get your Slack workspace ID. You can find it as follows:

    • From your desktop, open Slack in a web browser using its URL (e.g. app.slack.com).

    • Once the page loads and you sign in, the URL will be in the following format: https://app.slack.com/client/TXXXXXXX/CXXXXXXX.

    • Your workspace ID is the string beginning with T.

      Slack workspace ID highlighted in the browser URL

  2. Create a Slack app to access your workspace messages:

    • Visit this slack tutorial site to quickly create a Slack app and access token.
    • Click "Create App" and follow the install wizard. This creates a Slack application with some default scopes. (Note: for production use, edit the scopes to what you really need.)
    • Once the app is created, click "Install in workspace".
    • Click "OAuth & Permissions" and copy the "Bot User OAuth Access Token".
  3. Add a data store to your Agent Builder:

  4. Click Create data store, select Slack and enter your Slack authentication information. Click Authenticate.

  5. On the next screen, select what you want to sync. You can choose from Files, Messages and Conversations.

    Note: At the time of writing, Google Agent Builder can't sync private channels, multi-party instant messages or 1:1 instant messages.

    Slack Data Store Configuration

  6. Click continue and select a location for your data store. Enter a name for your data store.

  7. Select a synchronization frequency for your data and click Create. At the time of writing, there are three options:

    • Daily
    • Every 3 days
    • Every 5 days

Vertex AI now starts ingesting your Slack messages. To check the status of the ingestion, go to the Data stores page and click your data store's name.

Slack Data Store

Syncing data seems to take a while in general, so be patient.

Creating a Vertex AI Agent Builder search app

With the Slack data store set up and the data indexed, we can create an app that uses this data. Vertex AI Agent Builder offers 4 types of applications:

  • Search: a search application that uses the data store as well as web search to answer users' questions. This is the closest to "traditional" RAG.
  • Chat: a chatbot-like application where you define different chat paths. It's more like pre-LLM chatbots, enhanced with LLMs. Compared to "just" search, chat allows more complex interactions and guiding the user.
  • Recommendation: a recommendation application that uses the data store to recommend items.
  • Agent: this option is quite interesting, as it lets you define an LLM agent system with multiple agents just by describing, at a high level, what the agents should do and which steps they should take to reach their goals.

NOTE: At the time of writing, only Search and Agent seem to be compatible with external connectors like Slack.

For this example, we create a search application.

  1. Navigate to the Agent Builders Console and click "Create new app".

  2. Leave the default settings, add a name and select a region. Click "Continue".

    Create a new app

  3. On the next screen, select the data store you created earlier. Click "Create".

    Select Data Store during app creation

  4. Once the app is created, you can test it in the Preview tab of the application. As we built a "Search" app, the preview is a nice and simple search bar.

    Slack Search App Preview

Advanced LLM features of the Vertex AI search app

At first glance, the app is just a search bar that finds information. But it is enriched with Google's Gemini LLMs. They summarize search results and let users ask follow-up questions, which is enormously helpful in a search context.

On the "Configuration" page of the app, the "Advanced" option lets you enable or disable these LLM features.

Advanced Search App Configuration

I highly recommend keeping them enabled, but they come with a price tag:

  • Default search pricing: $2 per 1000 queries
  • With Enterprise features enabled: $4 per 1000 queries
  • With LLM features enabled: an additional $4 per 1000 queries

With all features enabled, you pay $8 per 1000 search queries.

Test the search behavior with and without these features to decide whether you need them.

Deploying a Vertex AI Agent Builder app

With the app up and running, we want to make it accessible to our users. Google offers 3 deployment options:

  1. Web app: a hosted page that can be accessed via a predefined URL. This is the easiest option, as it requires just a click on "Enable". The drawback: it's just the search input box, so it's not that useful for most people.

  2. Widget: a widget that you can embed in your website, which puts the search functionality right where your users are. It's especially useful for customer-facing websites.

  3. API: an API you can use to integrate the search functionality into your own applications. This is the most flexible option.

How to deploy the Vertex AI search app as a widget

To complete the example, here are the steps to integrate the Slack search app into your own website:

  1. Navigate to the Agent Builder Console

  2. Select the app you just created.

  3. Click "Integration" in the left-hand menu.

  4. Click "Widget". Enter your domain, click "Add" and then "Save".

  5. This creates a JS snippet similar to this one:

    1<!-- Widget JavaScript bundle --> <script
    2src="https://cloud.google.com/ai/gen-app-builder/client?hl=en_US"></script>
    3
    4<!-- Search widget element is not visible by default -->
    5<gen-search-widget
    6configId="<config-id>"
    7location="eu"
    8triggerId="searchWidgetTrigger">
    9</gen-search-widget>
    10
    11<!-- Element that opens the widget on click.-->
    12<input placeholder="Search here" id="searchWidgetTrigger" />

    Add this snippet to your website. It creates a hidden search widget that opens when a user clicks the input field. The triggering element doesn't have to be an input. It can be any HTML element, as long as its id matches the triggerId of the widget.

  6. Finally, create an OAuth token and pass it to the widget element.

    1// Set authorization token.
    2const searchWidget = document.querySelector("gen-search-widget")
    3searchWidget.authToken = "<JWT or OAuth token provided by your backend>"

An API integration works much the same way: select "API" instead of "Widget" in the integration tab.

Conclusion

In this post, we built a Google Search-like application on top of data in our own systems, in this case Slack, using three main elements of the Vertex AI Agent Builder: external connectors, data stores and search applications.

Everything was done in the UI, with no code at all, so you can build an application like this without a data scientist or AI expert. The external connectors are a low-code way to connect external systems and data sources. The data stores seem to be vector stores that you then connect to applications. The search application is easy to integrate into existing websites and also offers an API.

The limitation: while the Search app was easy to configure, it was, in the end, just search. Compared to modern RAG systems with chat capabilities, that was a bit disappointing. If you are comparing cloud options, our Azure AI Search RAG tutorial shows the Microsoft equivalent, and Copilot Studio offers a similar no-code route with SharePoint lists as knowledge sources.

The "Agent" and "Chat" apps promise to be much more flexible than the "Search" app and are worth a look if you need more than search.

Further reading

Interested in building high-quality AI agent systems?

We prepared a comprehensive guide based on cutting-edge research for how to build robust, reliable AI agent systems that actually work in production. This guide covers:

  • Understanding the 14 systematic failure modes in multi-agent systems
  • Evidence-based best practices for agent design
  • Structured communication protocols and verification mechanisms
Andreas Nigg

About the author

Andreas Nigg

Co-Founder, Pondhouse Data

Andreas has spent more than eight years building data applications. At Pondhouse Data he designs and ships AI systems and integrates them into complex, real-world environments.