We spoke with Sarthak Kumar Barik and Stephen Conneely about the engineering sector and how team dynamics are evolving with the times.

As technologies advance, siloed working environments have the potential to become a thing of the past, particularly as we find more convenient and effective ways to stay in contact with globally dispersed peers. 

The engineering space is no different and for Stephen Conneely, the director of QA engineering at Fidelity Investments Ireland, interdisciplinary engineering has reshaped how teams deliver results, especially in an environment where AI-assisted development has become more commonplace.  

Conneely told SiliconRepublic.com: “Teams bring together software engineers, quality engineers, analysts and platform specialists to jointly own problems end‑to‑end, with AI tools supporting activities such as test design, code review and documentation.

“This shared ownership reduces hand‑offs and allows risks to surface earlier, while maintaining strong governance and accountability. Quality is designed in from the start, with disciplines collaborating to decide where AI accelerates delivery and where human validation remains critical.”

He explained that this is all happening in an atmosphere where engineers are expected to understand how their work impacts adjacent systems, data integrity and the client experience. “The result is teams that move faster with confidence, using AI as an enabler rather than a shortcut and delivering more predictable outcomes in complex environments.”

This is echoed by Workhuman’s Sarthak Kumar Barik, a principal engineer, who stated: “As a platform team, our work does not exist in isolation. Product teams across the organisation own use cases built on the same legacy foundation and they face the same migration challenges, often without the same depth of context.” 

It is the responsibility of engineers and other employees, he finds, to close that gap in knowledge. He explained this can be achieved by translating the migration experience into reusable patterns, clear guidance and well-defined integration points that other teams can adopt without starting from scratch. 

He said: “We work alongside product teams as active partners, helping them map their existing behaviour to the new platform, identifying where the gaps are, and making sure each migration they undertake is faster and less risky than the one before.

“The goal is that knowledge compounds across the organisation rather than staying locked within a single team.”

A more tangible example of this reshaping, he explained, is in how the organisation uses artificial intelligence in the engineering workflow. Workhuman ran an AI-assisted workshop where developers provided context about the system, its architecture, data flows and constraints, and used this as the foundation for AI-generated code. 

“The difference compared to generic prompting was striking. When AI is given the real context of your system, it becomes a genuine accelerator, producing code that is relevant, grounded and faster to review and adapt. This has changed how both our team and the product teams we support think about velocity.”

Barik added: “Interdisciplinary engineering, for us, is less about organisational charts and more about shared context. When platform and product teams work from the same understanding of the system, the target and the tools available, progress accelerates across the board.”

With that in mind, what skills and processes do engineers need to be on top of to ensure they are keeping pace with change across the sector?

Fundamentals and the future

For Conneely, the most in-demand skills, in today’s engineering landscape, are a combination of the fundamentals alongside the adoption of emerging technologies.

He said: “We continue to prioritise deep capability in software engineering, quality engineering, cloud platforms and data, but increasingly value engineers who can use AI‑assisted tooling responsibly to improve productivity, quality and decision making.”

Engineers should also prioritise the ability to critically evaluate AI‑generated output, apply sound engineering judgement and develop the ability to understand where human oversight is essential, as well as adopt a systems thinking, automation-first approach and risk-based decision making, which Conneely said are as important as framework or language expertise. 

“Just as critical are communication skills, particularly in regulated environments where engineers must explain technical decisions, including AI usage, in clear business terms. As technology evolves, learning mindset and adaptability are now core competencies rather than ‘nice‑to‑haves’.”

Similarly, for Barik, the challenge is often in matching critical but older systems with newer, more advanced models and processes. He explained that the challenge is not just technical, but more intuitive, as you have to figure out whether you are actually making progress when the system is deeply coupled and cannot be taken offline.

He said: “We defined the target architecture upfront, not as an aspiration but as a concrete end state against which every decision is measured. From there, we decomposed the system into smaller subsystems with a roadmap of agreed milestones. Each milestone represents a discrete, verifiable unit of progress, a subsystem dialled down in the legacy platform and enabled in the new one. 

“Every pragmatic shortcut taken along the way is recorded as technical debt, so the team always knows exactly what remains rather than discovering it later. The most powerful measure of progress has been observability. By instrumenting both old and new systems, we track in real time what percentage of load is flowing through the new platform versus the legacy one. 

“A subsystem is not truly migrated until the traffic data confirms it. Progress is not a milestone ticked off – it is a measurable, visible shift in where the load is flowing.”

Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.