Seven Steps to DataOps

We believe that there are seven key steps that analytic teams must follow to implement DataOps.

Add Tests

Add Data And Logic Tests

Test your data and test logic/code/scripts used in each tool in your analytic toolchain.

End to End Process

Put All Steps To Version Control

At the end of the day, analytic models/workbooks/scripts is all just code.

Branch and Merge

Branch and Merge

Branching and merging enables people to safely work on their own tasks.

Multiple Environments

Use Multiple Environments

Your analytic work requires coordinating multiple tools and hardware.

reuse and containers

Reuse & Containerize

Reuse code and results. Manage the environment tool with Containers.

Parameterize

Parameterize Your Processing

Parameters save you time by allowing you to call subsets of process.

Focused on Analytic Teams

Orchestrate Two Journeys

Orchestrate the journey from data to value and the journey from idea to production

Learn More: Read the DataOps Whitepaper


Download Whitepaper

Overview

Our platform takes those Seven Steps of DataOps and implements them in software.

Kitchens Are Where You Work




kitchens
What Are Kitchens?
  • Kitchens allow people to create their own environment for analytics.
  • Team members make changes in their Kitchen and merge those back into production.
Why Kitchens Matter:
  • Kitchens let team members safely work together or separately.
  • Merging separate Kitchens is easy, quick, and safe.

Recipes Span The Analytic Process

What Are Recipes?
  • Recipes are steps in a directed graph stored as a set of text files.
  • Recipes encapsulate the complexity, do the orchestration work, and tests the results.
Why Recipes Matter:
  • Steps contain multiple tools, environments and technology options.
  • Step execution is distributed across a cluster of machines.

recipes

Built in Order History


orders
What Are Orders?
  • When every Recipe is executed Orders save the ‘as-run’ information.
  • Order Run saves status, test results, timing and other metadata.
Why Orders Matter:
  • Order Run history enables statistical process control.
  • Order Run history provides for data governance and reuse.

Ingredients Are Reusable

What Are Ingredients?
  • Kitchen, Recipes, and Orders allow analytic teams to work independently. However, teams need to work dependently as well. Ingredients allow one team’s work to be incorporated into another team's Recipe.
Why Ingredients Matter:
  • Allow reuse someone else’s processed data.
  • Easily incorporate summarized transforms or analytics.
  • Safe differentiation of production data vs test data.



ingredients

Cloud Native Architecture

architecture
What is the Architecture?
  • The DataKitchen platform is built from the bottom up with a multi-tenant and secure cloud design.
Why The Architecture Matters:
  • Every customer utilizes its own encrypted cloud: data, servers, databases, and software.

Platform Interfaces

What Are The Interfaces?
  • DataKitchen has three flexible ways to interact with the system: user interface, command line, and API.
Why Interfaces Matter:
  • Easy to use web-based user interface for non-technical users.
  • Rich command line interface for technical users.
  • Complete REST API integrates into existing systems.



interfaces