Deep Dive Recap: What’s Next for Engineering in the Age of AI with VPE Forum CEO Melanie Ziegler

For engineering leaders, the pace of change meant that 2025 felt less like a year and more like a decade. Now, at the start of a new year, they’re anticipating yet another eventful year for software engineering in the age of AI.

To tackle the question of “what comes next?” Melanie Ziegler, founder and CEO of the VPE Forum, joined Jellyfish advisor Adam Ferrari for our December Deep Dive webinar. VPE Forum is a peer groups organization that brings together Engineering VPs and CTOs to exchange practical insights, tackle complex challenges, and strengthen leadership. Member organizations that have benefited include CloudZero, Toast, and Whoop among many others, giving Ziegler a unique high-level view of the industry.

Their in-depth conversation covered everything from financial planning and hiring trends to context engineering and autonomous agents, with examples from across the VPE Forum community.

Here are some of the highlights.

From Experimentation to Expectation

AI adoption reached 90% in 2025 according to our 2025 State of Engineering Management report. That increase in adoption reflects the industry’s changing attitudes towards AI. At the start of the year, engineering leaders allowed early adopters to experiment with AI. Just a few months later, the same leaders expected everyone on their team to be trying the tools and sharing their experiences. We’re now at the stage where AI literacy is often mandatory for software development – from experimentation to expectation in under 12 months.

The next phase of AI adoption also comes with a behavioral shift. Teams aren’t just using AI tools to speed up coding – they’re turning to AI at every stage of the software development life cycle (SDLC), from talking to users and creating product requirements to reviews and deployments.

This year, we’ve seen AI go from novelty to real utility,” said Ziegler. “Software development has changed faster than any other abstraction shift, and in our VPE Forum groups, every quarterly meeting, every monthly sync, every community event includes discussion and conversations on AI.

Impact is About Business Results

There’s a growing body of evidence that AI is having a positive impact on business results. According to a joint study by Jellyfish and OpenAI, cycle times dropped by 24% when organizations moved from low to high day-to-day adoption. The 2025 State of AI-Assisted Software Development report from DORA also found a strong correlation between AI adoption and productivity.

The findings from these reports match what Ziegler has been hearing from VPE Forum members. AI is a force multiplier, allowing 30 or 40 developers to act like a team two or three times the size. Ramping up AI adoption also helps compress timelines – work that used to take weeks or quarters can now be completed in days.

One of our community members used to invest heavily in upfront customer research and UX design; now they’re able to very rapidly ship working software,” said Ziegler. “They achieved going from nothing to a real number of live customers and substantial revenue in a relatively small number of months. That’s what you call real impact.

Ziegler emphasized that impact is about business outcomes, not activity. As organizations increase their AI investments, being able to quantify how AI impacts delivery, quality, and productivity across the SDLC will be more important than ever.

The Promise of Agentic Development

Many engineering organizations are looking beyond augmenting their work with AI to a future in which autonomous agents can complete complex tasks from beginning to end. Some VPE Forum members are already experimenting with agentic AI, and they’re optimistic about its potential.

One of our members is using AI to draft, validate, and stress test requirements and tech plans, with the expectation that in 2026 those end-to-end tasks will be performed with only human review instead of any human authorship,” explained Ziegler. “The role of the human engineers will become more about supervising and steering.

Context engineering will be critical to achieving success with agentic AI. Just like human engineers, AI agents need access to all the relevant data and documentation in order to complete a complex task. However, Ziegler points out that most companies have work to do before they can start reaping the rewards.

Companies will need to make meaningful structural improvements to engage agentic workflows. Investing in consistency of tools and processes across teams will be a priority for many in 2026.

Optimism Around Hiring

2025 was a wait-and-see year, with companies slowing hiring in response to larger economic conditions and the uncertainty surrounding AI. There are signs that hiring will improve in 2026, driven in part by an acceptance that developers remain critical to delivering product.

Several companies across our community are hiring,” said Ziegler. “I’m also seeing instances of companies bringing interns back. That pipeline matters – interns are naturally going to grow into full-time hires. The conversation is back on the table, and that’s encouraging.

New Demands on Engineering Leaders

AI is reshaping the role of the engineering leader, easing workloads in some areas while also adding new responsibilities.

As engineering organizations move from experimentation to implementation, executive teams will hold leaders accountable for AI spend. Engineering leaders will be expected to work within a fixed budget for tool acquisition and token spend, and invest in a way that delivers measurable business results.

We’ve certainly seen how costs can spiral if they’re not managed,” said Ziegler. “I think there is going to be an increased focus on financial responsibility in 2026, and executive teams will be involved with AI spend in the same ways they are today with cloud spend.

AI: Power Tool for Engineering Leaders

AI isn’t just for technical teams; it’s a management power tool for engineering leaders. AI can act as a junior analyst, digging through engineering metrics and identifying anomalies or concerns. Beyond day-to-day tasks, AI can help leaders align strategy with business goals and monitor team performance.

Change management is critical to successful AI adoption. While AI coding tools can improve the developer experience, the pace of change can be exhausting. To keep teams engaged and avoid burnout, Ziegler recommends that engineering leaders:

• Lead with clarity and honesty
• Prioritize partnership
• Encourage experimentation and curiosity
• Celebrate wins
• Stay focused on outcomes

Ziegler also encourages leaders to talk to other leaders, to learn from each other, and share their successes, problems, and challenges. It’s a learning curve, and we’re all in it together.

We don’t need all the answers,” said Ziegler. “Stay curious, stay honest, focus on outcomes, experiment, and we’ll see where we are a year from now.

For more insights, watch the full webinar recording here. To find out more about measuring AI impact with Jellyfish, request a demo.