Maki Nage

Stream Processing Made Easy

Maki Nage is a Python stream processing library and framework. It provides expressive and extensible APIs. Maki Nage speeds up the development of stream applications. It can be used to process stream and batch data. More than that, it allows to develop an application with batch data, and deploy it as a Kafka micro-service.

Read the doc to learn more.

Main Features

• Expressive and Extensible APIs: Maki-Nage is based on ReactiveX.
• Deployment Ready: Maki-Nage uses Kafka to scale the workload, and be resilient to errors.
• Unifies Batch and Stream processing: The same APIs work on both sources of data.
• Flexible: Start working on your laptop, continue on a server, deploy on a cluster.
• ML Streaming Serving: Serve your machine learning model as a Kafka micro-service.

Installation

Maki Nage is available on PyPI:

pip install makinage

Getting started

Write your data transforms

import rx
import rxsci as rs

def rolling_mean():
    return rx.pipe(
        rs.data.roll(window=3, stride=3, pipeline=rx.pipe(
            rs.math.mean(reduce=True),
        )),
    )

Test your code on batch data

You can test your code from any python data or CSV file.

data = [1, 2, 3, 4, 5, 6, 7]

rx.from_(data).pipe(
    rs.state.with_memory_store(rx.pipe(
        rolling_mean(),
    )),
).subscribe(
    on_next=print
)

2.0
5.0

Deploy your code as a Kafka micro-service

To deploy the code, package it as a function:

def my_app(config, data):
    roll_mean = rx.from_(data).pipe(
        rs.state.with_memory_store(rx.pipe(
            rolling_mean(),
        )),
    )

    return roll_mean,

Create a configuration file:

application:
    name: my_app
kafka:
    endpoint: "localhost"
topics:
    - name: data
    - name: features
operators:
    compute_features:
        factory: my_app:my_app
        sources:
            - data
        sinks:
            - features

And start it!

makinage --config myconfig.yaml

Serve Machine Learning Models

Maki Nage contains a model serving tool. With it, serving a machine learning model in streaming mode just requires a configuration file:

application:
    name: my_model_serving
Kafka:
    endpoint: "localhost"
topics:
- name: data
  encoder: makinage.encoding.json
- name: model
  encoder: makinage.encoding.none
  start_from: last
- name: predict
  encoder: makinage.encoding.json
operators:
  serve:
    factory: makinage.serve:serve
    sources:
      - model
      - data
    sinks:
      - predict
config:
  serve: {}

And then serving the model it done the same way than any makinage application:

makinage --config config.serve.yaml

Some pre and post processing steps are possible if input features or predictions must be modified before/after the inference:

Read the book to learn more.

Publications

• Toward Data Science: Stream Processing Made Easy
• KDnuggets: Real-Time Histogram Plots on Unbounded Data

License

Maki Nage is publised under the MIT License.