How Will AI Transform DevOps?

Share on facebook
Share on twitter
Share on linkedin
Share on reddit

Run-through

The value of AI-powered Business Intelligence

What benefits can AI-powered Business Intelligence can bring? While this may be of interest to true blue BI Managers and Analysts, business intelligence is making its way into a lot of other roles. Between skill shortages, job title overlap, and a growing interest in cultivating cross-functional teams, BI is pretty much for everyone who has a mind to be more efficient, productive, and profitable.

Read More »

The Differences of DevOps and DataOps

How different are DevOps and DataOps? Are we splitting hairs or hairstyles? Both are aligned to Agile principles and have a number of processes in common, but they have some key differences, too. Both are also very complex and aim to solve mostly different sets of challenges while also working to improve team cross-functionality. So, let’s take a look to get a better understanding of both.

Read More »

What if you could push a button and magically achieve 100% efficiency across all teams, tasks, and projects instantaneously? That’s a question to ponder over the next 20+ years as we watch AI and DevOps transform almost everyone’s roles. Agile software development thrives when we have cross-functional teams. One of the goals of DevOps is to create increasingly cross-functional teams. And AI… well, AI makes achieving cross-functionality a lot faster, and arguably easier.

It’s a lot to take in to be sure. No one said that becoming Borg would be easy – but it might be far easier than we’ve imagined. Is that being facetious?

First, definitions:

DevOps entails practices, automated tools, and the cultivation of a team mindset to synthesize software development, quality assurance, and operations into increasingly cross-functional teams to increase the speed, efficiency, and quality of software development relative to both team and software business objectives. AI is a computer system that can perform tasks that previously required humans to do.

DevOps and Automation

DevOps can vary in maturity from one team or application to the next – as with the practical evolution of Continuous Integration, Delivery, and Deployment. Puppet’s 2016 State of DevOps indicated that teams practicing Continuous Delivery had 2,555x faster lead times, deployed 200x more frequently, had a 3x lower change failure rate, and 24x faster Mean Time to Recover Rate (MTTR) than teams doing everything manually. Bring on the automation! Puppet’s State of DevOps Reports for 2020 and 2021 are chock-full of insights, too. DevOps concerns the entire SDLC and arguably matters before the SDLC even starts. That’s a tremendous amount of ground to cover. Even so, AI makes it easy and fast for DevOps Teams to get data-driven answers with actionable insights into how they can improve any aspect of their teams’ development efforts.

How Can AI Help DevOps?

The AI-powered digital assistant we are developing in our Early Access Program will answer questions using behavioral analytics to suggest “next actions” based on best practices in Agile software development. It is important to recognize that AI’s role is that of an advisor. You and your fellow humans are the decision-makers. The AI references must be transparent so that if you have the inclination you can see the data and how it is calculated.

Digging through data is a time-consuming exercise that can be hit or miss in terms of generating actionable insights. Previously, we looked at how a software engineering manager could spend an hour examining a developer’s performance to come up with ways to reduce their defect rate. Each defect prevented (historical vs future trend) applies to their ROI as cost-savings.

An AI-powered Digital Assistant cuts through the time you spend digging through performance analytics to provide, in essence, “Insights on-demand.” What used to take hours can be achieved in minutes:

  1. Ask a question
  2. Get an answer
  3. See the data the answer is based on
  4. Receive a visual representation of the data
  5. Have a clear course of action of what to do next

The following are just a few of the DevOps questions that AI can help answer:

Direct Ways AI Can Help DevOps

Improve Productivity

  • What are each developers’ strengths and weaknesses?
  • Which developer is the best in what implementation?
  • Who delivers the most value?
  • Who generates the least waste?
  • Who manages their backlog with the best quality?

Improve Efficiency

  • What are each teams’ strengths and weaknesses (value, velocity, reduced waste, data quality)?
  • Are our backlogs prioritized and managed the right way?
  • Do team members touch anything that’s not part of the sprint?
  • How can I optimize the delivery process?

Budget Optimization

  • How much do I spend per team?
  • How much do I spend on waste?
  • How much do I spend on bug fixes vs new implementation vs enhancements?
  • To which teams should I allocate more budget to maximize the value of the deliverables?
  • How can I allocate the budget for a maximized ROI?

Here’s an example of how we envision it working:

Use Case: What are your developers working on?

The Data

  • Coding Hours: total estimated hours spent on writing code within a selected timeline
  • Code Volume: total lines of code inserted and 20% of line deleted in a commit
  • Distribution of coding hours and code volume between developers or repositories

Summary:

Based on the chart, the Bitcoin repository definitely has the greatest focus among the developers. While the Dogecoin repo has an irregular small amount of development time allocated which started to be more consistent since mid-May of this year.

The million-dollar question for a scenario like this: Does the allocated work align with your organization’s plans and focuses? Or is the team’s work allocation not prioritized correctly?

Action

Review your delivery targets and budget allocation, then verify if the current situation is disrupting any short or long-term goals. If you find any disconnection between your plan and the workload on different projects, try to clarify the tasks for your team and reorganize work allocation.

No Improvement? What to Do Next?

What if it’s not getting better?
Maybe your team is lacking specific technology skills and needs some training? Check the technology stack used for the project. A lack of experience in a particular language – as Python – might need time to research and self-teaching instead of dedicating effort to write new code.

Is DevOps or AI Better?

To simplify. DevOps is a cultural mindset and set of practices to improve stuff just as AI is a tool to automate stuff. Both can be implemented without the other, but they go together… like a Reese’s Peanut Buttercup!

Our view is that “Everything starts with the Team” – so, DevOps is it is embodied by people engaging to improve many things, including their own cross-functional capabilities. AI makes DevOps faster.

DevOps By Any Other Name...

Even Puppet’s tenth annual report for 2021 questions, “Do we know what DevOps is yet?” Obviously, they do, we do and you do, too – but not everyone does. How we define and describe it should vary because the acceptance of and maturity of its implementation varies from one organization to the next. Puppet also notes that the practice of DevOps is not limited to those who have DevOps in their job or team titles.

This brings us to the continuous drive to become more cross-functional inherently leads to more people becoming cognizant of and ultimately practicing DevOps regardless of their job or team title. What manager does not want their team to be more efficient and productive in delivering higher quality software? So, while DevOps may not be an explicit line item in every manager’s official duties, it’s implicit. Leastwise, we don’t necessarily need to be DevOps specialists or managers to be practicing DevOps.

That doesn’t change the nature of the task, it just means that many aspects of DevOps are relevant to everyone – at least, eventually. This is not to diminish the role and function of DevOps Specialists and Managers now for being part of the process to get everyone there.

DevOps and AI Beyond the SDLC - Human Resources?

Where does cross-functionality start and end? One huge impact on the SDLC is developer skill. Say it ain’t so. While we’ve talked about retention and turnover before, we haven’t really spent a whole lot of time on the hiring process – at least not outside of developing onboarding programs. But, hiring and recruiting, training, and long-term continuing education programs are huge, too.

With the Great Resignation going on alongside everyone no longer being forced to work from home, but now insisting on it – we’ve added a little bit of complexity to the equation. The one big mitigating factor in all of software development today, globally – is a major shortage of sufficiently skilled IT workers of all stripes. It’s important for all software development teams to make the best use of everyone on their team – and everyone they hire. Developer wages are typically the largest expense in all software development projects, so budgets are a thing (for most companies).

This is where DevOps and Human Resources intersects. Granted, not every company or team has an HR team or manager. Hiring processes can vary widely. Engineering managers may be the hiring authority. You might rely upon recruiters or employee referrals, job fairs and job boards. Many companies outsource their development, others use team augmentation – relying upon IT staffing companies.

A huge topic unto itself – but what if you could measure the effectiveness of every recruiting venue and every development agency? Stay tuned as we’ll take an in-depth look into this area, too.

About Gitential

Gitential is an Analytics and Engineering Intelligence service provider bringing visibility and optimization highlights on teams’ productivity.

Our mission is to augment decision-making with smart-recommendations to improve DevOps and DataOps teams’ performance and to proactively mitigate risk areas within development projects while bringing visibility and turning the development sector more fact-based.

Please take a moment to arrange a free demo – or sign up for a free trial, no credit card is needed.

Did you like our content?

Spread the word

Share on facebook
Share on twitter
Share on linkedin
Share on reddit

Subscribe to Our Newsletter

Don't miss our latest updates.
All About Software Engineering Best Practices, Productivity Measurement, Performance Analytics, Software Team Management and more.

Did you like our content?

Spread the word

Share on facebook
Share on twitter
Share on linkedin
Share on reddit

Subscribe to Our Newsletter

Don't miss our latest updates. All About Software Engineering Best Practices, Productivity Measurement, Performance Analytics, Software Team Management and more.