AIOps in the Value Stream: My 2019 Vision for Smarter DevOps
Including AI in your DevOps process makes it easier to see and influence the overall value stream
Join the DZone community and get the full member experience.Join For Free
It's that time of year when everyone looks into the crystal ball and casts predictions for the New Year. As a society, we expect smarter technology in the future. We envision a world where machines do all the work. Where are the flying cars we were promised?!
We want machines that do household chores, give us recommendations according to our likes and interests, and in general, take over tasks that humans don't like. Artificial Intelligence (AI) removes unpleasant or tedious jobs from our plates, and that is what makes it so appealing. AI frees humans up to focus on more creative work. Beyond that, AI allows for personalization — machines that know you and act according to data.
It's no different in IT: we utilize AI to make decisions faster and handle processes that follow specific rules. AI analyzes a set of data and then acts accordingly. We (technologists) don't want to have to search through massive sets of data to make business decisions — so we automate.
For example, if a financial organization just updated its online banking application and the software development management team needs to decide what features to focus on building next, those leaders can use AI to help with the decision-making process. These decisions are based on what features are most used and most popular, or based on negative feedback or service tickets related to a specific use of the application. Today, increasingly, that is happening through intelligent automation.
When decisions and actions stem from data history, we consider those to be intelligent decisions, hence the "I" in AI.
When I look ahead at 2019, I anticipate increased use of AI and analytics driving decisions within the software product value stream.
AI and Analytics Driving Decisions Within the Software Product Value Stream
In previous blogs and webinars, I discussed "making work visible" and how that work is linked to value. You see, Value Stream Management (VSM) is the easiest way to ensure software development and delivery is bringing optimal value to the business. To discover the value of certain workflows within the software development lifecycle, you need some form of measurement and monitoring.
By identifying waste, and focusing on delivering value based on business goals, Value Stream Management is the best way for a lean enterprise to succeed in today's software-driven world. Measuring efforts related to people, teams, tools and processes, as well as end-user feedback, are essential for making informed decisions and tweaking value streams.
The more data leveraged, the more intelligent the decision. Thanks to the adoption of Agile and DevOps, we see a growing trend where AI and automation are essential to creating and releasing great software products.
We also see AI being used to continually optimize the value stream. By using AI, we gain a better understanding of the value within our teams, tools, and processes. AI supports value stream thinking, rather than project-based thinking. Again, utilizing business priorities as a key ingredient to drive the value stream.
We also see great potential for AI in the areas of Lean and process improvement. AI is helping determine where we can pick up slack and cut more waste, and the tools leveraged within the tool chain that capture these metrics are invaluable to organizations.
Of course, APIs are critical for reporting and leveraging data for AI, and that will be one of CollabNet VersionOne's priorities early this year — intelligent integration and collaboration through technologies like webhooks will make exposing this work more visible and valuable for our customers. In addition, advanced analytics and measurement capabilities in the value stream will be integrated and leveraged with other solutions that support the business.
Having the ability to monitor value streams and act — based on specific metrics, defined by multiple stakeholders — is AI in action within DevOps.
There has been increasing discussion of "AIOps" in DevOps communities, and the emphasis here is on data analytics. Many are saying it's the next iteration of DevOps — one that relies on automation to make changes to the value stream, based on the measurement of activity. This approach results in less manual analysis of the value stream, which to my point at the beginning, is the ultimate purpose of AI — to take on manual tasks and to make fast, data-driven decisions that matter to the business!
I look forward to seeing how AI is used in the coming year to improve the way organizations plan and release software. I'm eager to see further connections between DevOps and AIOps. Most of all, I look forward to sharing new CollabNet VersionOne innovations we have in the works to ensure software delivery provides the most value for the business possible.
Published at DZone with permission of Eric Robertson. See the original article here.
Opinions expressed by DZone contributors are their own.