# Why Prism?

In the world of data management, orchestrators are a critical tool for building and operating data pipelines. One of the most popular orchestrators is Apache Airflow, thanks to its Python-based design and web interface. 

However, for many teams, Airflow may not be the best choice for supporting the entire process of creating and running data pipelines. Airflow's design originated in a time when software engineering principles weren't yet fully integrated into the world of data management. As a result, it may not fully meet the needs of modern data teams.

For example, here are a few differences between Airflow and Prism.

1. **Local development**\
   Airflow has an extremely abrasive local development environment. Users generally have to spin up a web server and a scheduler, which itself often requires knowledge of Docker, Redis and Celery. As a result, it's difficult to develop, test, or review pipelines outside a production deployment. Conversely, local development with Prism is easy. You can launch a run immediately using the [CLI](https://docs.runprism.com/cli) in a lightweight way.
2. **Data as a first-class citizen**\
   In order to move data between tasks. users can either write the data to S3 (or some other cloud platform) or use XCOMs. These have a ton of limitations (e.g., you have to write custom logic to serialize / deserialize data, large XCOMs cause the web server to crash, etc.). Prism doesn't share these limitations. Tasks can return any object, and the `CurrentRun.ref(...)` call will reference that object exactly without any additional logic needed.
3. **Imperative vs declarative**\
   Airflow's DAG are imperative: you have to keep track of the pipeline order and provide instructions for which task should run when. In contrast, Prism projects are declarative. The pipeline order is automatically inferred by looking at function calls.

This list is not exhaustive, and the full slate of differences between Prism and Airflow can fill its own blog post.

**TL;DR:** Prism was designed to make data practitioners more productive. We'd love for you to join our community!


---

# 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.runprism.com/getting-started/why-prism.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.