# Tips & Tricks

### Too many nodes?

When you work with many nodes, loops, or branches, nodes are always registered in the workflow as a flat list. Having too many nodes can make it difficult to visually follow the execution flow and focus on a certain part of the workflow for changes or new development.

Instead of a flat 20-node array, break it into logical sections:

```php
 class MyWorkflow extends Workflow
 {
     ... 
    protected function nodes(): array
    {
        return [
            ...$this->classificationProcess(),
            ...$this->analysis(),
            ...$this->postProcessing(),
        ];
    }
    
    /**
     * Initial classification process
     */
    protected function classificationProcess(): array
    {
        return [
            // Nodes here...
        ];
    }
}
```

The spread operator (...) keeps the runtime behavior identical. It's still a flat array. But now the
\
`nodes()` method reads like a table of contents, and each group is self-documenting.

## Example Projects

### AIForm - Conversational Data Collection

AIForm is a component for collecting structured data through multi-turn natural language conversations. It uses an AI agent to progressively gather information defined by a structured output class, validating each piece of data along the way.

The form maintains conversation history, tracks collected fields, missing fields, and validation errors across multiple turns.

<a href="https://github.com/neuron-core/ai-form" class="button secondary" data-icon="github">Check out the GitHub repository</a>

### Deep Research Agent

This project is inspired by Open Deep Research, which uses LangGraph for implementation. Other implementations exist also for llamaindex, and others. Our version leverages Neuron to create a powerful, modular workflow for research and analysis.

Neuron Open Deep Research provides a structured approach to generating comprehensive research reports on any topic using large language models, with a focus on modularity, extensibility, and real-time results.

#### Architecture

**DeepResearchAgent**: Orchestrates the overall report generation process

* **Planning**: Creates the structure of the report
* **GenerateSectionContent**: Generates content for each section using search results
* **Format**: Compiles the final report

**SearchWorkflow**: Handles search operations as a nested workflow

* **GenerateQueries**: Creates search queries based on section topics
* **SearchTheWeb**: Executes parallel searches and processes results

<a href="https://github.com/neuron-core/deep-research-agent" class="button secondary" data-icon="github">Check out the GitHub repository</a>

### Travel Planner Agent

This project demonstrates how to create a tour planner using Neuron PHP framework for agentic applications.

Stack Used:

* Neuron Workflow for multi-agent orchestration.
* [SerpAPI](https://serpapi.com/) for finding hotels, flights and places to visit comprehensive research reports on any topic using large language models, with a focus on modularity, extensibility, and real-time results.

#### Architecture

**TravelPlannerAgent**: Orchestrates the overall itinerary generation process

#### Nodes

* **Receptionist**: Collect all the information from the user
* **Delegator**: Generates single reports for flights, hotels, and places to visit
  * *Flights*
  * *Hotels*
  * *Places*
* **GenerateItinerary**: Generates the final report

<a href="https://github.com/neuron-core/travel-planner-agent" class="button secondary" data-icon="github">Check out the GitHub Repository</a>

### Laravel Travel Agent

This project demonstrates how to integrate multi-agent workflows in a Laravel application using Neuron PHP AI framework.&#x20;

Stack Used:

* [Laravel](https://laravel.com/) and [Livewire](https://livewire.laravel.com/) for the application.
* [Neuron Workflow](https://docs.neuron-ai.dev/workflow/getting-started) for multi-agent orchestration.
* [SerpAPI](https://serpapi.com/) for finding hotels, flights and places to visit comprehensive research reports on any topic using large language models, with a focus on modularity, extensibility, and real-time results.

<figure><img src="/files/2tRw8eZPhkZxW3Ms4i2i" alt=""><figcaption></figcaption></figure>

#### How to use this project

Download the project on your machine and open your terminal in the project directory. First, install the composer dependencies:

```bash
composer install

npm run build

php artisan migrate
```

Create a `.env` file in your project root (see `.env.example` for a template), and provides the API keys based on the service you want to connect with.

```
# At least one required
ANTHROPIC_API_KEY=
GEMINI_API_KEY=
OPENAI_API_KEY=

#Required
SERPAPI_KEY=

# Optional
INSPECTOR_INGESTION_KEY=
INSPECTOR_TRANSPORT=sync
```

Open the project in your browser, register an account, and start planning your trip.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.neuron-ai.dev/workflow/examples.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.