> For the complete documentation index, see [llms.txt](https://guides.noloco.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://guides.noloco.io/nola/getting-started-with-nola.md).

# Getting Started with Nola

Getting started with Nola is simple. This guide will show you how to access Nola and start using her to build your app.

## How to Access Nola

Nola is available directly within the Noloco app builder. Here's how to open Nola:

1. **Open your Noloco app** in Build Mode
2. **Look for the Nola tab** in the top-left tab of your app studio
3. **Click the icon** to open the Nola chat interface

That's it! Nola is now ready to help you build.

{% hint style="info" %}
If you don't see the Nola tab, make sure you're on a supported plan (Free, Build, or Enterprise). Some legacy plans don't have access to Nola.
{% endhint %}

## Your First Conversation with Nola

When you first open Nola, you'll see a friendly welcome message and some suggested actions to get started. Here are some great first requests to try:

### Example Starter Requests

**For New Apps:**

* "Help me create a Customers table with contact information"
* "Generate 10 sample records for my Projects table"
* "Set up a basic CRM structure for my business"

**For Existing Apps:**

* "Link my Customers table to my Orders table"
* "Create a [workflow](/workflows/workflows.md) that sends an email when a task is marked complete"
* "Add a [dashboard page](/pages/blank-pages.md) with [charts](/views/display/charts.md) showing my sales data"

## Understanding Nola's Responses

When you make a request, Nola will:

1. **Confirm she understands** what you're asking for
2. **Take action** or ask clarifying questions if needed
3. **Explain what she did** so you can learn and verify
4. **Suggest next steps** to continue building

### Example Conversation

**You:** "Create a Tasks table"

**Nola:** "I'll create a Tasks table for you. What fields would you like to include? I suggest:

* Task Name (text)
* Description (long text)
* Due Date (date)
* Status (single select: To Do, In Progress, Complete)
* Assigned To (linked to Users)

Would you like me to proceed with these fields, or would you like to customize them?"

**You:** "That looks perfect, go ahead"

**Nola:** "Done! I've created your Tasks table with those 5 fields. The table is now available in your data sources. Would you like me to:

* Create a view to display these tasks?
* Generate some sample data?
* Set up a workflow for task notifications?"

## Tips for Communicating with Nola

### Be Specific

The more details you provide, the better Nola can help.

❌ **Vague:** "Add a form" ✅ **Specific:** "Add a form for creating new customer records that includes name, email, phone, and company fields"

### Break Down Complex Requests

For complex tasks, consider breaking them into steps.

❌ **Too Complex:** "Build me a complete project management system with tasks, clients, timesheets, and invoicing" ✅ **Step by Step:**

1. "Create a Projects table with name, client, start date, and status"
2. "Now create a Tasks table linked to Projects"
3. "Add a timesheet tracking feature"

### Use Natural Language

You don't need special commands or syntax—just talk to Nola naturally.

✅ "Can you help me..." ✅ "I need to..." ✅ "How do I..." ✅ "Show me how to..."

### Ask Questions

Not sure how something works? Just ask!

✅ "How do [workflows](/workflows/workflows.md) work in Noloco?" ✅ "What's the difference between a [rollup](/data/collections/rollups.md) and a [lookup field](/data/collections/lookup-fields.md)?" ✅ "Can you explain [permissions](/users-and-permissions/user-roles-and-permissions.md) to me?"

## Attaching files and images

You can give Nola more than words. Use the attachment control in the chat to add files and images to a request:

* **Images and screenshots** — share a mockup, a screenshot of another tool, or a design for Nola to work from ("build a page that looks like this").
* **Documents and specs** — attach a brief or specification and ask Nola to build from it.
* **Data files** — upload a CSV, or a document such as a receipt, and Nola can create records from it. She'll ask you to approve the records before they're added.

## Using voice input

If you'd rather talk than type, use the voice input control in the chat to dictate your request. It's useful for longer, more detailed prompts — describe what you want out loud and Nola works from what you say.

## How Nola works through a request

For anything beyond a quick change, Nola works like a capable teammate rather than a single-shot command:

* **She plans first** — for a larger task, Nola lays out a short plan of the steps she'll take, so you can see her approach before she starts.
* **She shows her thinking** — as she works, Nola streams her reasoning and the steps she's taking, which you can expand or collapse.
* **She asks when it matters** — if a request could go a few ways, Nola asks a quick multiple-choice question rather than guessing. Pick an option, or skip it to let her decide.
* **She checks before big changes** — Nola asks for your approval before actions that change or remove things, so nothing significant happens without your say-so.

And once changes are made:

* **Changes are immediate** — you'll see them reflected in your app builder.
* **Changes can be undone** — use Noloco's undo button or manually revert changes.

{% hint style="warning" %}
Always review changes Nola makes, especially before [publishing](/settings/publishing.md) your app to your team or clients.
{% endhint %}

## Controlling what needs your approval

By default, Nola asks for your approval before actions that **change or remove** things, while read-only actions run without interrupting you. You decide where that line sits — both in the moment and up front.

### From the approval card

When Nola pauses for approval, you can **Approve** or **Deny** the action for that one time. To make your choice stick — so Nola stops asking for the same kind of action — expand **More approval options**:

* **Allow for this chat** — until the current conversation ends.
* **Allow "*****this table*****" only** (or a specific page or workflow) — every time that particular table, page, or workflow is used.
* **Always allow** — every time, anywhere in this app.

Anything you choose here is saved to your **Tool permissions** settings, so the two always stay in sync.

### From your Tool permissions settings

Nola's settings include a **Tool permissions** page where you can decide up front how much Nola can do on its own. These settings apply to you, in this app.

Actions are grouped into categories you can expand — **Records**, **Tables and fields**, **Pages**, **Action buttons**, **Workflows**, **App and settings**, and **Memory**. For each action you can choose:

* **Always ask** — Nola pauses for your approval every time.
* **Always allow** — Nola never pauses for it.

Until you set an action yourself, it follows **Nola's default** — asking before changes and removals, and running reads without a prompt. You can set an entire category in one go, and any per-table, per-page, or per-workflow exceptions you've granted from an approval card appear here as badges you can remove at any time.

This lets you give Nola a longer leash for routine work while still requiring a check on the things that matter most.

{% hint style="info" %}
Some larger jobs are carried out by a sub-agent — building page sections, editing workflows, and generating [Canvas](/nola/canvas.md) pages. This work runs without pausing, regardless of your Tool permissions settings.
{% endhint %}

<figure><img src="/files/kEzsU5V0a9uIO1r7AkHE" alt=""><figcaption><p>Nola's Tool permissions settings</p></figcaption></figure>

## Nola's Context

Nola is context-aware, meaning she understands:

* What app you're working on
* What page you're looking at
* What tables and fields you have
* Where you are in the builder
* What you've previously discussed in the conversation

This means you can have natural, flowing conversations without repeating context.

**Example:** **You:** "Create a Customers table" **Nola:** \[creates table] **You:** "Now add an Orders table linked to it" **Nola:** \[knows "it" refers to Customers and creates the link]

### Point Nola at a specific element

When you want Nola to focus on one part of a page, select that element in the builder. The selected element is shown in the chat as context — so instead of describing which piece you mean, you can select it and simply say what you want changed.

### Remembering across conversations

Within a chat, Nola remembers what you've discussed. Beyond a single chat, she also keeps a longer-term [memory](/nola/memory.md) of your preferences and app conventions, so she stays consistent from one conversation to the next.

## Working across conversations

Your work with Nola is organised into conversations, so you can keep separate threads for separate pieces of work:

* **Start a new conversation** for a fresh task, to keep its context focused.
* **Switch between conversations** from the conversation list and pick up where you left off — your history is kept.
* **Delete a conversation** you no longer need.
* **Edit and resend a previous message** to try a request a different way, without retyping it.

## When to Use Nola vs. Manual Building

### Use Nola When:

* Setting up new data structures
* Generating sample data
* Creating [workflows](/workflows/workflows.md)
* Learning new features
* You want to save time on repetitive tasks

### Build Manually When:

* Fine-tuning visual design
* Making small tweaks to existing elements
* You want precise control over every detail
* Nola can't *yet* do what you need her to

{% hint style="success" %}
**Best Practice**: Use Nola to get 80% of the way there quickly, then manually fine-tune the remaining 20% to your exact preferences.
{% endhint %}

## Getting Help

If you're not sure what Nola can do, just ask:

* "What can you help me with?"
* "Show me examples of what you can do"
* "Help me understand \[feature name]"

## Next Steps

Now that you know how to access and communicate with Nola, learn more about what she can do:

{% content-ref url="/pages/OZcYeVgozL69b8jLqLgq" %}
[What Nola Can Do](/nola/what-nola-can-do.md)
{% endcontent-ref %}

{% content-ref url="/pages/0647kJQ2WZkWSoFTDaXF" %}
[Tips and Best Practices](/nola/tips-and-best-practices.md)
{% endcontent-ref %}


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## 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, and the optional `goal` query parameter:

```
GET https://guides.noloco.io/nola/getting-started-with-nola.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

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.