Upgrade to Pro — share decks privately, control downloads, hide ads and more …

ROI of AI

ROI of AI

Most companies have some experimentation or mainline usage of AI in their engineering workflows, but how is the ROI of that AI usage being measured? How do we know it’s helping? What traits in engineering have changed? Join Michael Stahnke, a seasoned VP of Engineering, researcher, and author, as he unveils the metrics and strategies that demonstrate the value of AI in enhancing engineering outputs.

Drawing from years of experience managing teams of various sizes and formal research that underpinned Puppet’s State of DevOps Reports, Michael will provide a comprehensive framework for assessing AI’s impact on developer productivity. The talk will explore key metrics at both team and departmental levels, providing insights into the critical measurements necessary when evaluating your investment in AI. Attendees will learn robust methods to capture and analyze data, ensuring an accurate assessment of AI’s contributions to engineering efforts.

Mastering the art of calculating ROI will be a central theme, with guidance on balancing tangible and intangible benefits to present a compelling business case. Additionally, Michael will share effective communication strategies to help you articulate the value of AI initiatives to stakeholders, ensuring alignment and support across your organization.

Whether you’re considering adopting AI tools or seeking to maximize their impact, this session will equip you with the knowledge and tools to drive informed decisions and foster a culture of continuous improvement in your organization.

Originally given at DevOpsDays Kansas City 2025

Avatar for Michael Stahnke

Michael Stahnke

August 13, 2025
Tweet

More Decks by Michael Stahnke

Other Decks in Technology

Transcript

  1. Measuring the Impact: The ROI of AI in Engineering Work

    fl ows Michael Stahnke VP Engineering at Flox (flox.dev) @stahnma
  2. @stahnma Mental Model for Metrics Baseline DevOps Metrics I don’t

    feel elite. Move Quickly with Confidence Throughput Quality Normalization Fungibility Rollbacks (or forwards) Deployments per Week Deploys / Dev / Week Successful deploys Narrow vs Wide Time to 5th PR Changes Understood Start Status Questions People Metric Metric Metric Themes Eng Metrics Framework Talk: https://www.youtube.com/watch?v=A1SCGGbHgZ0
  3. @stahnma cyn·i·cism noun an inclination to believe that people are

    motivated purely by self-interest; skepticism
  4. @stahnma Prompt: Generate me a meme image with the theme

    of "the most effective use of AI is to make memes faster and in higher quality"
  5. @stahnma ‣ Writing code ‣ Testing ‣ Debugging ‣ Observability

    ‣ Understanding Code ‣ Software Architecture ‣ Security ‣ Work breakdown ‣ Prioritization, dependencies ‣ Dealing with ambiguity ‣ Reliability, delivery accountability ‣ Evaluating Tradeoffs ‣ Delivering Feedback ‣ Seeking and receiving feedback ‣ Effective communication ‣ Knowledge Sharing ‣ Teamwork ‣ Relationship building ‣ Handling disagreement ‣ Decision making ‣ Driving alignment ‣ Process thinking ‣ Facilitation ‣ Mentoring ‣ Business acumen ‣ Strategic work ‣ Product Thinking
  6. @stahnma ‣ Writing code ‣ Testing ‣ Debugging ‣ Observability

    ‣ Understanding Code ‣ Software Architecture ‣ Security ‣ Work breakdown ‣ Prioritization, dependencies ‣ Dealing with ambiguity ‣ Reliability, delivery accountability ‣ Evaluating Tradeoffs ‣ Delivering Feedback ‣ Seeking and receiving feedback ‣ Effective communication ‣ Knowledge Sharing ‣ Teamwork ‣ Relationship building ‣ Handling disagreement ‣ Decision making ‣ Driving alignment ‣ Process thinking ‣ Facilitation ‣ Mentoring ‣ Business acumen ‣ Strategic work ‣ Product Thinking
  7. @stahnma More senior-level work doesn’t require as much coding The

    more senior the engineer, the less they like AI code
  8. @stahnma I’ve never worked with tailwind and HTMX before, could

    you walk me through making a landing page?
  9. @stahnma If you’re using AI exactly like you did traditionally

    software engineering, you won’t save much time.
  10. @stahnma Summary • If you’re not using AI tools, you’re

    may get lapped by those who are. • If you’re only using them, you probably don’t know what you’re doing. • If you’re saving time, you’re probably recovering cost. • If you’re learning and leveling up, you’re doing really well.
  11. @stahnma CircleCI Career Matrix (Public) ROI Calculator References ROI of

    Platform Engineering Talk Metrics Framework Talk Speakerdeck Let 1,000 Flower Bloom This deck
  12. @stahnma Availability
 Cost for Downtime COGS
 Cost for Services in

    support of Gross Margin Security 
 Attach Rate and dollar value for compliance docs and security questionnaires Developer Productivity
 Hiring Tradeoffs, time to value, throughput Developer Sentiment Developer Adoption Platform ROI Talk: https://www.youtube.com/watch?v=iVrq0CcMx0E
  13. @stahnma Mental Model for Metrics Baseline DevOps Metrics I don’t

    feel elite. Move Quickly with Confidence Throughput Quality Normalization Fungibility Rollbacks (or forwards) Deployments per Week Deploys / Dev / Week Successful deploys Narrow vs Wide Time to 5th PR Changes Understood Start Status Questions People Metric Metric Metric Themes Eng Metrics Framework Talk: https://www.youtube.com/watch?v=A1SCGGbHgZ0