Climb the five steps of a continuous delivery maturity model


It establishes a process through which a developer’s changes to an application can be pushed to a code repository or container registry through automation. Ways you can improve your organization’s performance against DORA metrics to achieve faster and more agile deployments. Automated deployment to a test environment, for example, a deployment that is triggered by pushing code to the development branch. Verifying that models meet the predictive performance targets before they are deployed. Testing the prediction service by calling the service API with the expected inputs, and making sure that you get the response that you expect.

However, an upfront complete redesign of the entire system is not an attractive option for most organizations, which is why we have included this category in the maturity model. The journey that started with the Agile movement a decade ago is finally getting a strong foothold in the industry. Business leaders now have begun to embrace the fact that there is a new way of thinking about software development. IT can once again start pushing innovation instead of restraining it by expensive, slow, unpredictable and outdated processes. There are many ways to enter this new era and here we will describe a structured approach to attaining the best results.

  • With continuous integration, new code changes to an app are regularly built, tested, and merged into a shared repository.
  • However, an upfront complete redesign of the entire system is not an attractive option for most organizations, which is why we have included this category in the maturity model.
  • Manual deployment to a production environment after several successful runs of the pipeline on the pre-production environment.
  • For example, you need to verify that the packages that are required by the model are installed in the serving environment, and that the memory, compute, and accelerator resources that are available.
  • It includes capabilities such as real-time monitoring, telemetry, and analytics.
  • For example, you have a function that accepts a categorical data column and you encode the function as aone-hot feature.

At, builds are typically triggered from the source control system on each commit, tying a specific commit to a specific build. Tagging and versioning of builds is automated and the deployment process is standardized over all environments. Built artifacts or release packages are built only once and are designed to be able to be deployed in any environment. The standardized deployment process will also include a base for automated database deploys of the bulk of database changes, and scripted runtime configuration changes.

MLOps level 0: Manual process

It can help organizations identify initial actions that provide the most significant effect, while indicating which practices are essential, and which should be considered advanced or expert. By plotting where you and your team sit against each of the pillars, you can also identify any areas that need more investment to bring you up to par before you start progressing to the next stage. Finally, sharing a maturity model with business stakeholders will also help to set reasonable expectations and communicate the benefits derived from CI/CD without reaching expert levels. Once you have established the foundations, you can look towards automating the first stages of your pipeline by extending your automated tests and collaborating with operations teams on the creation of pre-production environments.

Cloud Source Repositories Private Git repository to store, manage, and track code. Database Migration Service Serverless, minimal downtime migrations to the cloud. Cloud Code IDE support to write, run, and debug Kubernetes applications. Kubernetes Applications Containerized apps with prebuilt deployment and unified billing. VMware Engine Fully managed, native VMware Cloud Foundation software stack.


Optimizer combines related scripts together into build layers and minifies them via UglifyJS . All changes (code, configuration, environments, etc.) triggers the feedback mechanisms. Technology that makes it simple to roll back and forth between database versions. Optimised for rapid feedback and visualisation of integration problems.

How is continuous delivery related to DevOps?

In this whitepaper we outline a continuous delivery maturity model model based on experiences working with leading edge organizations who are implementing GitOps delivery pipelines. While every organization is different, a number of common patterns have emerged. An optional additional component for level 1 ML pipeline automation is a feature store. A feature store is a centralized repository where you standardize the definition, storage, and access of features for training and serving. A feature store needs to provide an API for both high-throughput batch serving and low-latency real-time serving for the feature values, and to support both training and serving workloads.

gap analysis

Teams need to implement a proper database testing strategy to optimize results. Continuous delivery is for everyone, not just the DevOps elite. After making any javascript or css changes, optimize the project using RequireJS Optimizer.

MLOps level 2: CI/CD pipeline automation

A basic delivery pipeline is in place covering all the stages from source control to production. At this level the work with modularization will evolve into identifying and breaking out modules into components that are self-contained and separately deployed. At this stage it will also be natural to start migrating scattered and ad-hoc managed application and runtime configuration into version control and treat it as part of the application just like any other code. At this stage, DevOps teams — continuous delivery experts all adopt some form of DevOps structure — have fully automated a code build, integration and delivery pipeline. They’ve also automated the infrastructure deployment, likely on containers and public cloud infrastructure, although VMs are also viable. Hyper-automation enables code to rapidly pass through unit, integration and functional testing, sometimes within an hour; it is how these CD masters can push several releases a day if necessary.

With a mature component based architecture, where every component is a self-contained releasable unit with business value, you can achieve small and frequent releases and extremely short release cycles. For a rapid and reliable update of the pipelines in production, you need a robust automated CI/CD system. This automated CI/CD system lets your data scientists rapidly explore new ideas around feature engineering, model architecture, and hyperparameters. They can implement these ideas and automatically build, test, and deploy the new pipeline components to the target environment.

At this level reporting is typically done manually and on-demand by individuals. Interesting metrics can e.g. be cycle-time, delivery time, number of releases, number of emergency fixes, number of incidents, number of features per release, bugs found during integration test etc. When moving to beginner level you will naturally start to investigate ways of gradually automating the existing manual integration testing for faster feedback and more comprehensive regression tests. For accurate testing the component should be deployed and tested in a production like environment with all necessary dependencies. At expert level some organizations choose to make a bigger effort and form complete cross functional teams that can be completely autonomous. With extremely short cycle time and a mature delivery pipeline, such organizations have the confidence to adopt a strict roll-forward only strategy to production failures.

Intelligent Operations Tools for easily optimizing performance, security, and cost. Network Service Tiers Cloud network options based on performance, availability, and cost. Network Intelligence Center Network monitoring, verification, and optimization platform. Network Connectivity Center Connectivity management to help simplify and scale networks. Cloud Load Balancing Service for distributing traffic across applications and regions. Cloud CDN Content delivery network for serving web and video content.

The next level in the continuous delivery maturity model entails defining the activities for the entire move-to-production process, along with the file and system locations plus tooling to automate it. The goal is to increase release cycles’ consistency, not their speed, although the intermediate stage is typically when organizations can stick to regular releases on a defined schedule, such as nightly or weekly. The entire CD process should be automated, launched with a single command. NISI has recently released the Continuous Delivery 3.0 maturity model, or CD3M. The Maturity Model guides the improvements of Continuous Delivery pipelines and/or software development processes in software organizations.

Agile transformation in heavy industries: An interview with SOCAR … – McKinsey

Agile transformation in heavy industries: An interview with SOCAR ….

Posted: Tue, 21 Mar 2023 00:00:00 GMT [source]

For your security, if you’re on a public computer and have finished using your Red Hat services, please be sure to log out. For more reference architectures, diagrams, and best practices, explore theCloud Architecture Center. Avoid having similar features that have different definitions by maintaining features and their related metadata. Components can have their own version of the runtime environment, and have different languages and libraries.


Andreas Rehn is an Enterprise Architect and a strong advocate for Continuous Delivery, DevOps, Agile and Lean methods in systems development. At beginner level, you start to measure the process and track the metrics for a better understanding of where improvement is needed and if the expected results from improvements are obtained. A typical organization will have, at base level, started to prioritize work in backlogs, have some process defined which is rudimentarily documented and developers are practicing frequent commits into version control. The purpose of the maturity model is to highlight these five essential categories, and to give you an understanding of how mature your company is.

Storage Transfer Service Data transfers from online and on-premises sources to Cloud Storage. Transcoder API Convert video files and package them for optimized delivery. Private Catalog Service catalog for admins managing internal enterprise solutions. Intelligent Management Tools for easily managing performance, security, and cost.

Previous Post
Newer Post

Leave A Comment