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

# Deployment - Python

> Deploy your Kinetix ML pipeline to Python in less than 5 minutes.

## Install the Library

Kinetix ML should be installed as a node module using the command below.

```bash theme={null}
pip install KMLPipePy
```

## Initialization

First start by importing the module.

```Python theme={null}
from KMLPipePy import KMLPipeline
```

Create the pipeline object using your project name, version, and API Key.

```Python theme={null}
pipe = KMLPipeline("[Your Project Name]", 1, "[Your API Key]")
```

Initialize the pipeline.

```Python theme={null}
pipe.initialize()
```

## Execution

To execute the pipeline you need to run the execute method with the required input data.

```Python theme={null}
pipe.execute([]) # the [] should be replaced with your list of inputs
```

## Webcam Inference

The pipeline will need a data source for each of its inputs. A common input is image data for nodes such as PoseDetection2D. For these nodes you can use OpenCV.

```Python theme={null}
from KMLPipePy import KMLPipeline
from KMLPipePy.types import Canvas
import cv2
import time

pipe = KMLPipeline("[Project Name]", 1, "[API Key]")
pipe.initialize() # initialize the pipeline

out = Canvas() # initialize output drawing canvas
cam = cv2.VideoCapture(0) # start OpenCV webcam capture

while True:
  res, image = cam.read() # read image from webcam

  if image is not None and image.any():
    out.set_image(image)
    t0 = time.time()
    outputs = pipe.execute([image, out]) # execute pipeline
    print(outputs)
    t1 = time.time()
    print(f"{1/(t1-t0)} fps")

    if out.show(1):
      break

cam.release()
```