Machine Learning Platform

Quix is a machine learning infrastructure platform with flexible microservice components that helps you build continuously deployable data-driven systems quickly.

To build a custom end-to-end platform in minutes, simply:

Create a workspace and streaming infrastructure

Create a workspace and add a new topic:

step 1

Choose whether to store or delete data flowing into that topic:

step 2

Organise data from different sources

Create additional topics to group and monitor any data, events or messages from any model, service or device:

step 5

Send data to the workspace

Use our C#, Python or HTTP SDK to stream telemetry, events and meta-data.

With our sample code, you can:

  • Send data directly from your applications or your devices
  • Host recorders on external servers to create a bridge between your data source and Quix, or
  • Deploy services into Quix to fetch or receive data

Try our simple Hello World example!

step 3

Build simple analytics features

Now that you are collecting good quality data you can analyse it and build insights into your frontend or a business intelligence tool using our Query API:

step 4

Connect your historic data and external services

Use topics and our SDK to connect securely to any data source or any external service. Persist this data to the Catalogue or call it on demand.

step 6

Develop, deploy and monitor models

Now that you have set-up your infrastructure, your data teams will be empowered to extract and deliver value. With Quix, your data teams can:

  • Build data models
  • Train them on historic data
  • Test them on live data
  • Deploy them to production
  • Monitor them throughout the product lifecycle

step 7

Develop and deploy custom software services

With your data team firing on all cylinders, your product and design teams can now create and deliver amazing experiences tailored to individual customers. Use Develop and Deploy to build:

  • Data collection services (recorders for existing hardware endpoints, document data scrapers, web crawlers)
  • Data-driven services which act on model output to send events and notifications to other models, services or product frontends.
  • Prediction & recommendation services that call ML models to fulfil a client request.
  • Customisation services which tailor experiences and content for each user.

step 8

There you have it: a fully featured data-driven production platform that the world leading innovators would be proud of!