# GPT

{% hint style="info" %}
:warning: This task is currently only available in the on-premise and open-source editions but not in the cloud version
{% endhint %}

"GPT" is a task that allows you to enrich or transform your data with AI using OpenAI's [chat completion](https://platform.openai.com/docs/guides/chat-completions) models.

### Usage

For example, you have such a pipeline consisting of two tasks:

<figure><img src="/files/5d1l2uBwpdUClgUjgosE" alt=""><figcaption></figcaption></figure>

The first task returns a JSON with the list of items, containing an item ID and amount:

```json
[
    {
        "id": 1,
        "amount": 33
    },
    {
        "id": 2,
        "amount": 150
    },
    {
        "id": 3,
        "amount": 4000
    }
]
```

Now you can ask chat GPT to transform this data. For example, multiply all amounts by 3:

<figure><img src="/files/yjs9VSUwMIC6tXxNxlOx" alt=""><figcaption></figcaption></figure>

And this is what the task returns:

<figure><img src="/files/OJ9WJUygD03Q92njHwp9" alt=""><figcaption></figcaption></figure>

Of course, such a simple multiplication can be done by other tasks without the use of AI, but that's just an example and you can use ChatGPT for much more. Some other ideas:

* Generate missing postal codes for the list of addresses
* Clean up organization names in the list
* Convert amounts from one currency to another
* Rewrite texts on the fly
* And so on and so on

For example, if you have a list of addresses missing postal codes like this:

```json
[
  {
    "country": "United States",
    "state": "Michigan",
    "city": "Detroit",
    "address": "2645 Woodward Ave"
  },
  {
    "country": "Canada",
    "province": "Alberta",
    "city": "Edmonton",
    "address": "104 Ave NW"
  },
  {
    "country": "United States",
    "state": "Florida",
    "city": "Sunrise",
    "address": "1 Panther Pkwy"
  }
]
```

With the workflow as above and that message to chat GPT:

<figure><img src="/files/MX7JaO4oVasY4GbH4Olk" alt=""><figcaption></figcaption></figure>

You will get the perfect result;

<figure><img src="/files/E2mDibNL0RLA1UBcAuxL" alt=""><figcaption></figcaption></figure>

### Configuration

To be able to use this task, you need to configure your API Secret Key and choose a model you want to use:

```bash
TASK_RUNNER_OPENAI_GPT_KEY=%%YOUR_KEY_IS_HERE%%
TASK_RUNNER_OPENAI_MODEL=gpt-4o-mini
```

The following two variables need to be placed in the file `./processor/taskrunner/.env`.

Read more about [API Keys](https://platform.openai.com/api-keys) and [available models](https://platform.openai.com/docs/models) in the official OpenAI reference guide.


---

# 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.ylem.co/pipelines/tasks-ip/gpt.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.