> ## Documentation Index
> Fetch the complete documentation index at: https://docs.flowx.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Agent Builder

> Create custom AI agents without code using the FlowX Agent Builder - a no-code/low-code platform for business users.

The FlowX Agent Builder is a no-code/low-code platform that enables business users to create custom AI agents without relying on IT or data science teams.

```mermaid theme={"dark"}
flowchart LR
    subgraph sources [" "]
        KB[Knowledge Base]
        API[APIs]
        DB[Database]
    end

    sources --> BUILD

    subgraph BUILD [Agent Builder]
        C[Configure] --> T[Test] --> D[Deploy]
    end

    D --> Chat
    D --> Workflows
    D --> API2[REST API]
```

Agents connect to your data through multiple integration options:

* **Knowledge Bases** - Ground responses in your documents using RAG
* **MCP servers** - Connect external tools via Model Context Protocol
* **REST APIs** - Call any HTTP endpoint
* **FlowX Database** - Query and store structured data
* **Kafka** - Event-driven integrations

## What can you build?

Agent Builder allows you to create specialized AI agents that can:

> [card] Process documents
>
> Extract, validate, and analyze data from various document types

> [card] Understand content
>
> Perform sentiment analysis, topic modeling, and named entity recognition

> [card] Generate content
>
> Create summaries, translations, and complete documents

> [card] Analyze images
>
> Recognize objects, extract text, and understand scenes

> [card] Make decisions
>
> Provide risk assessments and compliance checks with full auditability

## Key benefits

### Business empowerment

* **Drag-and-drop interface** accessible to non-technical users
* **Pre-built templates** for common use cases
* **Reduced time-to-market** for AI solutions

### Enterprise-grade features

* **Full audit trails** and explainable AI decisions
* **Integration with existing workflows** and systems
* **Compliance-ready** with regulatory requirements
* **Tool call limits** — agents are capped at 20 tool calls per run to prevent runaway operations, and at 10 calls to workflow tools within that run. Past the workflow-tool cap the call is not dispatched and the agent is asked to answer with what it already has, so the run continues rather than failing 

## Core agent categories

> [accordion] Document processing agents
>
> * Document extraction and validation
> * Cross-document consistency checks
> * Metadata creation and classification

> [accordion] Analysis agents
>
> * Sentiment and emotion analysis
> * Risk assessment and fraud detection
> * Customer behavior analysis for upselling

> [accordion] Generation agents
>
> * Report writing and document completion
> * Content summarization and paraphrasing
> * Multi-language translation

## Getting started

**1. Choose your approach**

* Start with **pre-built templates** for common scenarios
* Create **from scratch** with custom instructions

**2. Configure your agent**

* Define input sources (documents, images, structured data)
* Set processing logic and validation rules
* Configure output formats and destinations

**3. Test and deploy**

* Test with sample documents and data
* Review confidence scores and validation results
* Deploy into your FlowX workflows

## Agent Designer interface

The Agent Designer provides a visual canvas for building AI workflows:

| Component | Description |
| - | - |
| **Left panel** | Available actions and node types to drag onto canvas |
| **Canvas** | Visual workflow designer where you connect nodes |
| **Properties** | Configuration panel for selected nodes |
| **Test panel** | Run tests with sample data |

## Related resources

> [card] [Node types](https://docs.flowx.ai/6.x/ai-platform/agent-builder/node-types.md)
>
> Learn about Generation, Understanding, Extraction, and Analysis nodes

> [card] [Use cases](https://docs.flowx.ai/6.x/ai-platform/agent-builder/use-cases.md)
>
> Real-world examples: loan processing, mortgage validation, customer intelligence

> [card] [Developer Agentic Harness](https://docs.flowx.ai/6.x/ai-platform/developer-agentic-harness.md)
>
> Build your app in the Designer by describing changes to an AI agent

> [card] [Using agents](https://docs.flowx.ai/6.x/ai-platform/using-agents/overview.md)
>
> Learn how to integrate agents into your apps

> [card] [Chat-driven workflows](https://docs.flowx.ai/6.x/ai-platform/chat-driven-workflows.md)
>
> Build multi-turn chat experiences with session memory and intent routing


## Related topics

- [Agent Builder use cases](https://docs.flowx.ai/6.x/ai-platform/agent-builder/use-cases.md)
- [Intent Classification Agent](https://docs.flowx.ai/6.x/ai-platform/agent-builder/intent-classification.md)
- [Agent catalog](https://docs.flowx.ai/6.x/ai-platform/agent-catalog.md)
- [Agent typologies](https://docs.flowx.ai/6.x/ai-platform/patterns/agent-typologies.md)
- [Using AI agents](https://docs.flowx.ai/6.x/ai-platform/using-agents/overview.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.