AI in Software Delivery: It's About Your Team

Recently, I had a chance to review Google’s 2025 DORA Report, which assesses the current state of AI-assisted software development.

Read it here: DORA 2025 State of AI-Assisted Software Development

What struck me was the key takeaway from the authors:

“The research reveals a critical truth: AI’s primary role in software development is that of an amplifier. It magnifies the strengths of high-performing organizations and the dysfunctions of struggling ones.”

That single statement captures what I’ve seen firsthand as my teams have adopted AI tools.

AI Doesn’t Fix Delivery - It Amplifies It

At the end of the day, teams of people deliver software. AI tools can help, but if your processes, environments, or team culture are broken, AI will only highlight those weaknesses faster.

If all you do is focus on tools that speed up coding, the results will fall short.

Take Cursor, for example — a modern coding assistant that combines an AI editor and a chatbot to help software teams write, understand, and improve code faster. It’s an excellent tool for any delivery team. But if your environments are unstable or your processes are inefficient, the benefits you’ll see from Cursor will be limited.

What the 2025 DORA Report Reveals About Teams

The report identifies seven clusters of software delivery teams, each reflecting different levels of maturity and performance. Here’s a quick overview:

1. Foundational Challenges: Teams stuck in survival mode — low performance, weak processes, unstable environments.
2. The Legacy Bottleneck: Similar to the first, but burdened by aging systems that slow progress.
3. Constrained by Process: Stable operations, but inefficient workflows limit business value.
4. High Impact, Low Cadence: Producing strong work, but instability in operations reduces throughput.
5. Stable and Methodical: Excellent output and reliable systems, though delivery speed lags higher clusters.
6. Pragmatic Performers: Consistently deliver great work with strong velocity, but risks team burnout.
7. Harmonious High Achievers: The gold standard: high quality, high business value, and high team morale.

The Takeaway: Elevate the System, Not Just the Tools

The closer your teams are to Cluster 7, the greater the impact AI will have on your software delivery lifecycle. But AI won’t move you there on its own, leadership, process improvement, and healthy environments are what unlock AI’s full potential.

All software delivery teams will face the challenge of incorporating AI into their workflows. To see meaningful results, you must also understand where your teams stand today and work intentionally to improve the foundations of delivery alongside adopting AI.