The Future of DevOps

DZone 's Guide to

The Future of DevOps

The future holds more focus on security and DevSecOps, using AI/ML to improve processes, and more automation.

· DevOps Zone ·
Free Resource

To understand the current and future state of DevOps, we spoke to 40 IT executives from 37 organizations. We asked them, "What’s the future of DevOps from your perspective, where do the greatest opportunities lie?" Here’s what they said:


  • Security is huge. The more we automate the more we can automate problems. How to integrate security into DevSecOps. More connected = more exposure. What happens when next wave of tools comes out. What is going to be the driving tools after containers?
  • DevSecOps - security and privacy. Don’t need to know the details. Know the kind of information that is private. Common security things and privacy things you need to worry about. As a group working with educators here’s what you need to teach on security and privacy in addition technical. Human aspect cannot be ignored. More teaching on the socio part in addition to the technical side. See more collaboration. There will be a huge backlash with more disclosures coming out. 
  • Next two to five years merging operational stuff. Public cloud is solving a lot of issues. Opportunity to get more discipline about security. Finding a way to make that easier. 
  • As DevOps grows as the standard way of working across dependent IT teams, along with its subsequent rapid increase in global adaptation, one can expect more focus on DevOps in the fields of security, IoT and cloud computing. These are relatively nascent streams of DevOps right now, but the spread of DevOps into various uncharted fields of software could mean widespread implementation into various fields of technology and operations in the future.


  • DevOps with cloud-native and microservices allows you to revolutionize the app lifecycle – testing and production are integrated, and you see problems before you go live thanks to testing and troubleshooting. Analytics across the DevOps pipeline and runtime you can do things in a much better way agility with insight and control. Apply ML to determine risk. Bandwidth use, performance, see problems able to see code and configuration changes. Connecting the two together is a big thing. Can shorten the lifecycle with higher quality. Insight and control with security. 
  • More people are taking the leap from DevOps to data ops. As streaming architectures become more popular people will want and need more data ops and data logistics in their organizations because of ML adoption. Maturation of DevOps with the data becomes more pervasive. Simplifying the toolsets to pick up GPUs and run ML is becoming more pervasive and less scary. Use cases deliver high value. This is creating the launching pad for practical applications of big data in business. 
  • I see a lot of opportunities related to artificial intelligence (AI) and machine learning (ML) as it relates to DevOps. Anticipating how changes I make to the application affect the overall user experience. What do I need to test? What is the impact of a change? There is a lot of opportunities to help companies decide what to do next with their product as well as verify the impact of those changes, so they can confidently release high-quality software targeted at their end users’ needs. For example, with AI and ML, I can analyze hundreds of thousands of test results on my platform and determine those changes to a particular portion of the code base result in, on average, a 20% failure rate. The actions I can then take care to ensure that, whenever changes occur in this part of the code, we test much more thoroughly. An even better approach would be to refactor this portion of the code base. AI and ML can allow organizations to make better decisions faster by providing unique analysis correlating, code, test results, user behavior, and production quality and performance. This type of capability can have a dramatic effect on increasing the velocity and quality of DevOps pipelines. 
  • AI integration. To get good data you have to be able to repeat the process over and over again without human intervention. Repeatable and quantifiable processes. DevOps is about scaling out the same test over and over again. With AI it’s automated and the data reports automatically. AI development with DevOps. 
  • AI/ML data-driven understanding patterns are working and automating the findings of the patterns, remediation, going into a self-driving, self-healing. DevOps has to be built into that. This is my business environment, how do I build an app to reduce friction. Infrastructure needs to be set up based on the business the developer and the user.


  • Automated cloud-based testing. Security is integrated into DevOps upfront for DevSecOps. 
  • For big companies, which is who we focus on, renew your thinking, your processes, your architecture, and your applications. All of these make up your legacy—not just the applications. All need to be modernized. By renewing your thinking, I mean, embrace that you are turning into an IT company no matter what your business is. You’re no longer just a bank or an airline. Your software is central to the services you provide your customers. Since you’re now an IT company, you need to think about how to improve your IT processes so that they’re more streamlined, efficient, and producing value. Automation of application development and provisioning is a big part of efficiency and turning out better software. Also, look to reduce your number of applications, and use principles of modern software development to rewrite them. 80% of your code is old, never used, and is full of bugs. You need to modernize it to increase quality. 
  • DevOps is one area where they are watching the market and talking to start-ups. Testing, deploy management, process, security can all benefit from automation. 
  • 1) Opportunity to pursue what DevOps is supposed to mean. Pursue automation. Opportunity to mingle changes in automation with the cultural change to maximize flow, feedback, and continuous improvement throughout the process so problems are solved in the process. 2) Think about security more heavily. 3) Making work fulfilling and rewarding. 
  • We’ve seen more standardization and common tooling. DevOps is becoming more precise and scientific practice. More reference cases. Learning and tooling will spread to the community. More understood as a practice, more standardize and more precise and scientific. Opportunity to do more with merging DevOps with security. Huge potential with automation and CI/CD with DevSecOps and the value of automation in the security area.
  • 100% automation of CI/CD, unit testing, QA. More tools and expertise get better the more quickly you can run with your plans. Fully automated to continuously release software, features and bug fixes.


  • People on a DevOps team and looking at that as a quality of life factor when looking for a job. This will result in the coolest innovation for the companies. Focus on attracting the right talent and breaking down the silos.
  • Platforms to enable collaboration and automation. Once present in an organization enables greater success with predictability. Rather than failing fast all the time prevent failure by learning from your failures so you are failing less frequently. Tools, process, metrics, reporting take advantage of learning for continuous improvement. 10,000 releases per day at Amazon. Financial services want to reduce app dev time from 300 days to 120. Ability to articulate the value to get continued resource support. Gartner report when starting on DevOps, commit for three years. After one year, it's worse than when started. You need to break through the wall to see the benefits. Accrue analytics necessary to learn and improve.
  • People continuing to break down silos. People are getting smarter about bringing everyone to the table. As people bring about new projects they will think about building and deploying before thinking about rearchitecting. Growing pragmatism. Ask where we get to with intelligent build automation before we go cloud or microservice.
  • Infrastructure as code. A set of tools will be used for deployment. Ultimately it will be cloud providers handling deployment effectively removing the operation side.
  • If we expand the scope of thinking a focus on user stories it’s not a question of technology it’s a question of low-level infrastructure and a full spectrum of application delivery – observability, PaaS, alerting and dashboarding. Beyond operations to entire IT organization.
  • Exploring extensively. Right now it begins when writing the code ends. Writing the code is not part of DevOps thinking. We’re changing that since code generators are becoming popular. In integration generating the code is part of the DevOps pipeline. We’re exploring extending the DevOps pipeline to generating the code as well. The industry will collapse into two or three main product ecosystems – probably a dominant in every vertical a SaaS best practice.
  • The network. Need to envision to support the application for better time to service. Hasn’t changed much in 15 years. The network has not been designed to react to the needs of the application and the change in deployment. From the operators' perspective, there’s not much focus on day two operations and repeatable processes and the operation state of the environment at the moment.  Tooling is happening but its early in its infancy.
  • For DevOps to work, developers must be willing to expand their expertise and responsibilities. DevOps relies on a “many hands make light work” ethos which requires everyone to handle a piece of the development, operations, and security pie. This shouldn’t be viewed as a burden. DevOps is an opportunity for developers. DevOps enables companies to clearly define roles and responsibilities for developers which gives them autonomy and responsibility for the development of their project.
  • Today there are a lot of independent Lego blocks. K8s independent of cloud, CI is independent of CD, security, all are important and all are tiny dots. Will see a more holistic, integrated POV, from doing it yourself to integrated and validated as part of the flow. Simplification with alliances, acquisitions to make it easier.
  • We think the greatest opportunities lie in implementing and scaling DevOps in large enterprises. Enterprise IT is so diverse across companies and industries, so turning these horses into unicorns will have an incredible impact on the world. 
  • 1) In the future, DevOps will be popular in the Enterprise. Dev and Ops teams are increasingly moving to rapid development and reliable, high-performance services delivery. 2) Software enterprises will push towards a microservices infrastructure. Microservices allow organizations to architect an enterprise solution, independently, over a set of smaller services.
  • The biggest opportunity for DevOps is to drive into tech stacks and organizations that think that a move to DevOps requires a complete re-architecture of their application or adoption of replacement technologies. While those sorts of changes may be excellent opportunities to introduce culture change as well, teams running existing business-critical apps in “monolithic” architectures can take advantage of DevOps as well, if they choose the right tools.
  • I think containers continue to provide interesting opportunities as the tooling around them improves. Depending on how they’re used, they can confer a lot of benefits or a lot of downsides within a system. There are a lot of untapped or just-scratched ideas on how to use them to convey significant benefits, though. For a few examples of what I mean: using containers to sandbox applications for security and resource reasons. It seems basic on its face, but we mostly see containers used for stateless (and some stateful) server-side applications that are otherwise operated in traditional manners within those containers. The container is treated more like a packaging system than a runtime. What if you generated an application container per user or per user session? It provides a limitless array of opportunities for improving system security, user security, and adding additional analytics around user behavior or preventing resource contention between users of your application. As the tooling improves, containers become even lower-cost to deploy. I think there’s many yet to be tapped opportunities in how containers could be used by both technically-focused and traditional industry verticals to improve their business.
  • Platform-as-a-service is definitely a growing field. In the long-run, an application developer should be able to simply define a couple of entry points in their application package and that should be sufficient for the application to be deployed. I think this is very much aligned with what we have been doing in developing an abstraction between Applications and Technology Infrastructure. Another example is our Data Science Platform as a Service, where the DevOps of the platform itself is fully encapsulated away from the application developers who use this platform to develop machine learning powered applications.
  • Kubernetes (K8s) and Docker seem worthwhile to learn about for any DevOps team. Another one is scaling of machine learning. Not everyone is doing this and it’s new so there’s a lot to learn there. We’ve been diving into what it’s like to scale TensorFlow serving and I think that this is going to be a big part of DevOps in the future.
  • Containerization opens up a lot of possibilities in terms of revamped architecture. The surrounding architecture is isolated from everything else. This provides deployment mobility and flexibility in terms of what libraries are in use and what resources they need which is exciting. Another one is dealing with machine learning and IoT data. There’s a massive influx of data from devices and so we have to think about how we store it in a usable format and also the global distribution of it. People need to be able to get data from a variety of locations which brings politics into the mix. There are rules in development from countries all over the world that make determinations about how data is stored. It’s becoming kind of tricky to have international customers because data has to be stored differently in various physical locations. Finally, the move towards the SRE model. Operators will no longer be a separate team but work on each project with developers as they go. Part of this will be because of containers because you’ll be working together to build at the development phase. Containers couple the development process with the deployment process.
  • In today’s DevOps environments, many technology professionals have mastered operating containerized workloads at scale, and leveraging containers in production. DevOps teams that have mastered containers, and have begun deploying microservices, are swiftly moving into the future of DevOps. The future of DevOps will include increased deployment of microservices and implementation of a service mesh architecture to manage and monitor said microservices. Moving forward, tech pros will increasingly need to develop a good understanding of the breadth/depth of service mesh necessary for their environment and choose the right tool to fit their needs. The future of DevOps also lies in focusing on serverless computing and determining the use cases where serverless computing is appropriate for certain distributed environment and use cases. With this acquired skill, DevOps teams will be able to lead their organizations to strategic decision-making and ultimately into the age of digital transformation.
  • The future of DevOps is bright. I believe there are great opportunities to deliver purpose-built pipelines that would connect to common repositories and perform a set of necessary steps to reliably deliver working applications to known environments. All cloud-based of course.
  • Containers will continue to grow prevalence and influence within the modern technology landscape. DevOps has always been about automation and this technology allows for companies to more easily build and deploy infrastructure. More companies will adopt containers in an effort to improve automation capabilities and velocity. Organizations will also continue to seek out technologies that integrate into frameworks like K8s as work to further improve automation and velocity.

Here's who shared their insights with us:

ai, automation, devops, devsecops, enterprise devops, machine learning, security

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}