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

# Create a new chart

Building a new chart in Nebuly takes just a few clicks, and results arrive in seconds. Let's see how.

**1. Create a new chart**

Click the **Add new** button in the left sidebar and select **Chart** to open the chart builder.

**2. Select a metric**

Select the metric you want to analyze. You can find the full list of available metrics in the [Key Metrics](https://docs.nebuly.com/documentation/key-metrics.md) guide.

![Alt Text](https://mintcdn.com/nebulyai/h9aKiGKVR2DLe52O/images/analytics_variables.png?fit=max\&auto=format\&n=h9aKiGKVR2DLe52O\&q=85\&s=662e2e6ed0191a4db61a0496f577936e)

**3. Apply filters**

Use filters to narrow down your data and exclude what's not relevant. You can apply one or more filters at the same time. Nebuly supports 100+ filters — here are some examples.

| Filter | Description |
| - | - |
| **Topic** | The main theme the user is talking about during the interaction |
| **Data source** | The external reference or knowledge sources the LLM might use to fetch or validate information |
| **Language** | The specific language in which the model generates its response |
| **User** | The specific user ID |
| **Action** | The task or question the user performs during an interaction |
| **Frustration reason** | Specific reasons a user might be frustrated, such as verbosity, insufficient knowledge, lack of clarity, or the answer not being deep enough |

![Alt Text](https://mintcdn.com/nebulyai/h9aKiGKVR2DLe52O/images/analytics_filters.png?fit=max\&auto=format\&n=h9aKiGKVR2DLe52O\&q=85\&s=e0c6d92d6e4c2ce92829594e5dd5e40e)

**4. Group your data**

Use `Group by` to segment your data and compare different slices side by side. Like filters, Nebuly supports 100+ dimensions — here are some examples.

| Group | Description |
| - | - |
| **Topic** | The main theme the user is talking about during the interaction |
| **Data source** | The external reference or knowledge sources the LLM might use to fetch or validate information |
| **Language** | The specific language in which the model generates its response |
| **User** | The specific user ID |
| **Action** | The task or question the user performs during an interaction |
| **Frustration reason** | Specific reasons a user might be frustrated, such as verbosity, insufficient knowledge, lack of clarity, or the answer not being deep enough |

![Alt Text](https://mintcdn.com/nebulyai/h9aKiGKVR2DLe52O/images/analytics_groups.png?fit=max\&auto=format\&n=h9aKiGKVR2DLe52O\&q=85\&s=2b7b41ccbec680c5e99b9e0a6aec2435)

**5. Choose a visualization**

Select the chart type that best represents your data. While a `Line chart` is great for spotting trends over time, other types like `Horizontal bar chart` are more useful for comparing aggregated values across categories.

![Alt Text](https://mintcdn.com/nebulyai/h9aKiGKVR2DLe52O/images/analytics_charts.png?fit=max\&auto=format\&n=h9aKiGKVR2DLe52O\&q=85\&s=184292fc9f1fddc198dc7ade84e96792)

**6. Analyze your results**

Your chart is ready. You can add it to a report, or keep exploring by applying additional filters and group-bys to dig deeper into the data.


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