AI is rapidly transforming how software gets built – but the real story goes far beyond faster output. As teams accelerate, the work itself is changing, new bottlenecks are emerging, and the role of engineering is evolving in ways that aren’t always obvious at first.
To navigate this shift, engineering leaders are realizing that speed alone isn’t enough – you need clear visibility into what’s actually changing, where the pressure points are, and how to stay in control as the system around you evolves.
“We’re now seeing roughly 80% of our code generated by AI and up to 30% faster issue cycle times. But the numbers only tell part of the story – AI also introduced new bottlenecks we hadn’t anticipated, like CI pipelines under pressure from double the code volume. The real shift is that our engineers are no longer writing code hands-on; they’re doing system design, prompt engineering, and architecture work. It’s made engineering more important, not less.
What made all of this manageable was having visibility into it. When a transformation is moving this fast, operating blind isn’t an option. Jellyfish gave us the data to understand where we were seeing real ROI, where the new constraints were emerging, and how to stay in control of something that’s genuinely changing in front of our eyes.”
Ron Ben Yosef
VP, Technology and Business Operations | Loadsmart
The message is familiar to anyone who has managed a complex system: when throughput increases, pressure concentrates elsewhere. More code can mean more bottlenecks downstream like more review time, more integration work, and heavier continuous integration demand. Without instrumentation, leaders risk “winning” on speed while quietly accumulating failure downstream.
Julia Gan, engineering leader at Box, frames the challenge less as a single metric and more as an organizational requirement: understanding where AI is being used, how it differs across teams, and what “responsible” adoption looks like in practice:
“Jellyfish helped us uncover deeper insights into AI usage, adoption, and impact across our engineering teams. We now have a clearer understanding of velocity and capacity, and can see how different AI tools affect different teams. This visibility allows us to benchmark our practices against industry standards and demonstrate to key stakeholders how Box is leveraging AI responsibly and productively.”
Julia Gan
Sr. Director, Technical Program Management and Engineering Chief of Staff | Box
For engineering leaders, this is the crux: AI adoption isn’t a single decision, it’s a portfolio of behaviors. Different teams use different tools, at different intensities, with different outcomes. The management task becomes comparative and continuous – less “did we adopt AI?” and more “where is it helping, where is it distorting, and where is it creating new queues?”
A third account emphasizes what often determines whether AI becomes a durable lever: the ability to tie the change to business outcomes, not just developer anecdotes. For this engineering leader, visibility is what turns AI from “promising” to provable:
“Jellyfish has allowed us to prove AI’s strategic value to the business. After integrating Augment, our Jellyfish data showed a 50% decrease in issue cycle time and a 2x increase in both deployment rates and Epics resolved per month. Having that level of visibility transformed AI from an experimental tool into a proven engine for business growth. Our teams are shipping code faster and delivering twice the value to our customers in half the time.”
Tom Osowski
Engineering Manager, Partnerships | Taskrabbit
Taken together, these perspectives point to a clear pattern: AI doesn’t just change how engineering work gets done – it changes how it needs to be managed. Leaders who recognize AI as a shift in the system, rather than another new tool, will be better equipped to sustain progress without running into new bottlenecks down the line.
See how engineering leaders are driving engineering excellence in the age of AI with more success stories from LastPass and Flo Health here.
About the author
Adlana Estephanian is the Senior Customer Marketing Manager at Jellyfish.
