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

# Architecture Deep Dive

> Understand the core architecture and components of Julep

## Overview

Think of Julep as a platform that combines both client-side and server-side components to help you build advanced AI agents. Here's a mental model to help you understand it:

1. Your Application Code
   * Use the Julep SDK in your application to define agents, tasks, and workflows
   * SDK provides functions and classes for easy setup and management of components

2. Julep Backend Service
   * SDK communicates with Julep backend over the network
   * Backend handles task execution, session state, document storage, and workflow orchestration

3. Integration with Tools and APIs
   * Integrate external tools and services within your workflows
   * Backend facilitates integrations for web searches, database access, third-party API calls, and more

To help you visualize this, here's a diagram:

<Frame caption="Julep Architecture Overview">
  <img src="https://mintcdn.com/julep/Q4dHkhmm-IyLB_6v/images/julep-architecture.svg?fit=max&auto=format&n=Q4dHkhmm-IyLB_6v&q=85&s=cf5ead97f1da43438b01faedcfc87c35" width={1000} height={1000} data-path="images/julep-architecture.svg" />
</Frame>

<Frame caption="System Architecture">
  <img src="https://mintcdn.com/julep/Q4dHkhmm-IyLB_6v/images/architecture-diagram.svg?fit=max&auto=format&n=Q4dHkhmm-IyLB_6v&q=85&s=1830e9e6b66bb224b3e26bafcdacb5fc" width={1000} height={600} data-path="images/architecture-diagram.svg" />
</Frame>

## Core Components

<CardGroup cols={2}>
  <Card title="Agent" icon="robot">
    <p>The main orchestrator of your application, backed by foundation models like GPT4 or Claude.</p>
  </Card>

  <Card title="User" icon="user">
    <p>Users can be associated with sessions and are used to scope memories formed by agents.</p>
  </Card>

  <Card title="Session" icon="messages">
    <p>The main workhorse for Julep apps.</p>
  </Card>

  <Card title="Tool" icon="wrench">
    <p>Programmatic interfaces that foundation models can "call" with inputs.</p>
  </Card>

  <Card title="Doc" icon="file-lines">
    <p>Collections of text snippets indexed into a vector database.</p>
  </Card>

  <Card title="Task & Execution" icon="list-check">
    <p>Github Actions-style workflows for complex operations.</p>
  </Card>

  <Card title="Secret" icon="key">
    <p>Encrypted key-value pairs for securely storing sensitive information like API keys and credentials.</p>
  </Card>
</CardGroup>

## Infrastructure Components

<CardGroup cols={4}>
  <Card title="Document Store" icon="database">
    <p>The document store provides a vector database for semantic search along with a document management system. It supports various file types and handles automatic indexing and retrieval of documents.</p>
  </Card>

  <Card title="Task Execution Engine" icon="gears">
    <p>The task execution engine handles distributed task processing, state management, error handling with retries, and supports parallel execution of tasks.</p>
  </Card>

  <Card title="Secrets Store" icon="key">
    <p>The secrets store provides encrypted storage for sensitive information like API keys, credentials, and tokens. It uses AES-256 encryption and developer-scoped access controls to securely manage confidential data.</p>
  </Card>

  <Card title="API Layer" icon="code">
    <p>The API layer provides RESTful API endpoints, SDK support, authentication and authorization capabilities, and handles rate limiting and quotas.</p>
  </Card>
</CardGroup>

## Data Flow

<Frame caption="Data Flow Overview">
  <img src="https://mintcdn.com/julep/Q4dHkhmm-IyLB_6v/images/flow.svg?fit=max&auto=format&n=Q4dHkhmm-IyLB_6v&q=85&s=00760fdf464ea040d9e42482b5411b71" width={1000} height={1000} data-path="images/flow.svg" />
</Frame>

<Info>
  <p>The data flow in Julep is divided into three main components, each handling specific aspects of request processing and response generation.</p>
</Info>

<Steps>
  <Step title="Client Interaction" icon="arrow-right-to-bracket">
    * SDK or API calls initiate requests
    * Authentication and validation
    * Request routing
  </Step>

  <Step title="Task Processing" icon="gears">
    * Task decomposition
    * Step execution
    * State management
    * Tool integration
  </Step>

  <Step title="Response Handling" icon="arrow-right-from-bracket">
    * Result aggregation
    * Error handling
    * Client notification
    * State persistence
  </Step>
</Steps>

## Security

<CardGroup cols={3}>
  <Card title="Authentication" icon="lock">
    <ol>
      <li>**1.** API key-based access</li>
      <li>**2.** Role-based permissions</li>
      <li>**3.** Session management</li>
      <li>**4.** Token validation</li>
    </ol>
  </Card>

  <Card title="Data Protection" icon="shield">
    <ol>
      <li>**1.** Encryption at rest</li>
      <li>**2.** Secure communication</li>
      <li>**3.** Data isolation</li>
      <li>**4.** Access controls</li>
      <li>**5.** Encrypted secrets</li>
    </ol>
  </Card>

  <Card title="Monitoring" icon="chart-line">
    <ol>
      <li>**1.** Audit logging</li>
      <li>**2.** Performance metrics</li>
      <li>**3.** Error tracking</li>
      <li>**4.** Usage analytics</li>
    </ol>
  </Card>
</CardGroup>

## Conclusion

The modular design enables you to customize and extend functionality as needed, while the security-first approach ensures your data and operations remain protected. Whether you're building a simple chatbot or a complex AI workflow, this architecture provides the foundation you need.

## Support

If you need help with further questions in Julep:

* Join our [Discord community](https://discord.com/invite/JTSBGRZrzj)
* Check the [GitHub repository](https://github.com/julep-ai/julep)
* Contact support at [hey@julep.ai](mailto:hey@julep.ai)