# Python
Learn how to build and deploy Python application and run DB migrations.
With Buddy, you can create a pipeline that builds, tests, and deploys Python applications on a push to Git. The configuration is super simple and takes 10 to 15 minutes.
Configuration is very easy and takes only a couple of minutes.
## 1. Create a new project
Go to **Projects** and click **Create new project**. Select your Git hosting provider (Buddy, GitHub, Bitbucket, GitLab, or Private Server) and enter your project name.
## 2. Add a new pipeline
Navigate to **Pipelines** and click **New pipeline**. In the form, configure the following:
- **Name** - Enter a descriptive name for your pipeline (e.g., "Deploy to production")
- **ID** - Optional identifier for the pipeline
- **CPU** - Select architecture: **x64** or **ARM64**
- **Resources per pipeline run** - Choose resource allocation (default: Workspace default)
- **[Definition Location](/docs/pipelines/introduction/advanced-settings.md#definition-location)** - Optionally check "Store configuration in a Git repository" to use YAML files
Click **Add pipeline** to create it.
## 3. Configure workflow: contexts and triggers
After creating the pipeline, go to the **Workflow** tab to configure:
* **Contexts** - Click `+` to add Git branch/tag context or Environment. See [Contexts](/docs/pipelines/introduction/contexts.md).
* **Triggers** - Click `+` to add Git events, Schedule, or Webhook. See [Workflow Configuration](/docs/pipelines/introduction/creating-pipeline/workflow-configuration.md).
## 4. Add actions
Buddy lets you choose from dozens of predefined actions. In this example, we'll add 4 actions that will perform the following tasks:
* Build and test your Python application
* Upload application to server
* Run db migrations & restart application
* Send notification to Slack
### 4.1 Build and test your Python app
Build actions in Buddy are run in isolated containers run from official Docker images. When the pipeline is run, Buddy pulls the container, runs build commands, and uploads the results to the pipeline filesystem.
Look up and click **Python** on the action list to add it to the pipeline:
The Run tab lets you determine the commands to execute. The default commands are:
```bash
pip install -r requirements.txt
nosetests
```
#### 4.1.2 Python version
You can change the version of Python and install missing packages & tools in the runtime environment tab:
If your tests require a database to run, you can [attach it in the Services](/docs/actions/services-and-databases/predefined.md) tab:
### 4.2 Deploy application to server
The compiled application needs to be uploaded to the server. Head to the **Deploy** tab and select your upload action (Transfer in our case, where you can choose SFTP Target):
When adding the action you can choose what and where should be uploaded:
[Targets](/docs/targets/introduction.md) allow to save pre-configured transfer server settings to be used across multiple actions.
### 4.3 Run db migrations & restart application
Once the app is deployed, you can run additional commands on your server with the **SSH** action:
Enter the commands to execute in Run CMDs and configure authentication details in the Target tab:
### 4.4 Send notification to Slack
You can configure Buddy to send your team a message after the deployment. In this example we'll use **Slack**:
If you add this action in the On Failure tab, Buddy will only send the message if something goes wrong with your build or deployment.
## 5. Summary
**Congratulations!** You have just automated your entire delivery process. Make a push to the selected branch and watch Buddy fetch, build, and deploy your project. With Continuous Delivery applied, you can now focus on what's really important: developing awesome apps! 🔥
Bear in mind that this article is only a brief example of what Buddy can do. You can create additional pipelines for staging and production environments, integrate with your favorite services \(AWS, Google, Azure\), trigger tests on pull requests, build Docker images, and push them to the registry—the possibilities are unlimited.
If you want us to create a delivery pipeline for your project, drop a line to [support@buddy.works](mailto:support@buddy.works) – we'll be happy to help!
---
Original source: https://buddy.works/docs/basics/workflows/python