# Setting up your personalized project

**Note:** You must create and map **at least one variable** before you can publish the project.

This step also ensures that your base video, script, and data inputs are fully synchronized for accurate personalization at scale. Double-check all configurations during this stage, as they directly impact the quality and consistency of your final generated videos.

#### Step-by-Step Guide

**Step 1: Navigate to the Personalized Projects Section**

* Log in to your Gan.AI account.
* Click on the "Home" or "My Videos" page from the left sidebar and select the "Create New" option in the top corner
* Click “Create New Personalized Project” to begin.

<figure><img src="/files/I6CcCSfj4xtE4gPCmrD8" alt=""><figcaption></figcaption></figure>

**Step 2: Upload/Record Your Base Video**

* Choose the base video you recorded or record a new video.
* Rename the project from the top left corner if required.

<figure><img src="/files/Je9M8aPktWVwigQKjtOG" alt=""><figcaption></figcaption></figure>

* Ensure it meets your campaign requirements (see Tips for Recording & Setup article).

<figure><img src="/files/uNEvcGeTa68uIFhxQO0h" alt=""><figcaption></figcaption></figure>

**Step 3: Select the checkbox**

Once the video is uploaded, please select the checkboxes as requested and continue with the video personalization.

<figure><img src="/files/OXUCYVpX7rA849albtPm" alt=""><figcaption></figcaption></figure>

**Step 4: Assigning Voice AI variables**

* Click on a word or select multiple words where you want to apply video personalization.
* Enter the name of the variable you are creating (Tip: to personalize a client's name, enter the variable name as 'name'), then click on Save.

<figure><img src="/files/SEQrAXTAgvSfZ8c0RH2i" alt=""><figcaption></figcaption></figure>

After selecting the Voice AI variables, please click on Proceed and submit the video for model training.

**Once your video is submitted, the Gan.AI team will begin the model training process. You’ll receive an email notification as soon as your personalized video model is trained and the final project is ready to use.**

#### **Best Practices**

* Use variable names that are intuitive and match your data source (**e.g., @FirstName, @City**).
* Avoid placing too many variables in a single sentence — this could impact layout readability.
* Ensure your base video and background visuals are clean and consistent so personalization appears natural.
* Please ensure that the video you upload includes all final edits and changes, as Gan.AI does not trim or edit videos once they are submitted.
* Before sending your video for personalization, make sure you’re completely satisfied with the final version.
* Trim your videos before model training — trimming is not possible after training is complete.
* If your campaign videos are color corrected, ensure that your training videos are color corrected in the same way for best results.
* Keep variable text short to prevent overflow issues in on-screen elements.
* Maintain consistent lighting and framing in your base video to avoid visual mismatches during variable replacements.
* Use high-quality audio in your base recording to ensure accurate voice cloning and cleaner output.
* Test your variables with sample data before full generation to verify lip-sync, timing, and placement.

**Pro Tip: If your voice is not detected in the script, we recommend re-recording your video to ensure better clarity and accuracy. This helps the system correctly capture and process your voice.**

**For the best results, speak clearly and minimize background noise during recording.**


---

# 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://help.gan.ai/getting-started-onboarding/personalized-project/setting-up-your-personalized-project.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.
