CSIA podcast: Why AI-Native Workflows Will Change Everything

AI Is Not Just a Faster Interface

“The biggest misconception is that AI is primarily about chatbots and productivity assistance,” Zein says. “Most companies still approach AI as an add-on. They basically add a chatbot to a 25-year-old workflow and then declare the AI revolution complete.”
The problem with that approach is not that chatbots are useless. The problem is that it leaves the underlying workflow untouched. For decades, electrical engineers have conceptualised solutions in their heads, then manually transferred that thinking into schematics, device tags, cross-references, terminal plans, bills of materials, and cabinet layouts. That sequence – mental model to manual documentation – has been left untouched for decades. The real question AI forces onto the table is not how to accelerate those steps, but whether they need to exist at all in their current form. As Zein puts it: “The real opportunity is not in helping engineers draw faster. The real opportunity is fundamentally changing how engineering itself is created.”

For more context on this shift, WSCAD’s article Why AI is redefining the logic of electrical design maps out the transition from AI-assisted to AI-native engineering in three phases.

“CAD, as we traditionally know it today, will probably become the execution layer, while AI becomes the decision layer,” Zein explains. “And that’s what we mean by AI-native engineering.”
Engineering value shifts upstream – from drafting fidelity toward the judgement calls that AI cannot and should not make autonomously. This is not, Zein emphasises, an incremental software update. “We are not talking about a chat window inside a CAD system,” he says. “We are talking about a new engineering paradigm.”

Further reading: AI Native Engineering is more than just another Copilot

Which Electrical Engineering Workflows Will AI Transform First?

Zein is specific about where the impact lands first: the repetitive, documentation-heavy tasks that consume a disproportionate share of engineering time without requiring the judgement that defines the profession. Every electrical project, without exception, involves the same recurring overhead:

• Generating bills of materials from schematic data
• Creating terminal diagrams and terminal plans
• Interpreting requirements from customers or upstream departments
• Cross-referencing components and managing numbering schemes
• Validation passes and error checking across documentation sets
• Propagating design changes consistently across an entire project

“Engineering today is still highly sophisticated copy-paste management,” he says. “Many engineers spend enormous amounts of time not actually engineering, but managing complexity.” A discipline that demands years of training has evolved, in many organisations, into a system for managing the bureaucratic overhead that the engineering process itself generates.

WSCAD’s 2024 global survey of 1,267 electrical engineers across 40 countries puts a number to this: 54% reported they do not have time for innovation. They are too focused on delivering projects on time and on budget. That is not a workforce motivation problem. It is a structural capacity problem – and AI is, in Zein’s view, the only credible path to solving it. The goal, he says, is workflows that become “more automated, more system-driven, and much more knowledge-centric – not so much about me having to perform a specific function in the system.”

Practical reference: How to use the ELECTRIX AI Copilot – how text commands trigger real engineering actions in ELECTRIX AI.

Will AI Replace Electrical Engineers?

It is the question in every room when the topic of AI comes up. Zein answers it directly: no, electrical engineers will not disappear. But the role will change more significantly than most people currently expect.
Every major technological leap, he argues, changed the role of the people involved rather than eliminating them. “Probably future engineers will look back at some of the current workflows the way we look at manually drafting with ink pens,” Zein says. The directional shift is clear: less time drawing, more time deciding – defining systems, validating outcomes, orchestrating automation. The engineer’s authority over technical decisions does not diminish. The mechanics of producing documentation stop consuming it.
Zein uses a concrete historical analogy to illustrate the point. The transition from horse carriage to railway did not primarily eliminate coachmen – it replaced a job with an ecosystem. Most individual coachmen did not make a clean transition to train conductor. But one train created more roles than a single coach ever could, and society ended up with far more trains than it ever had coaches. Job expansion followed displacement, at a scale the previous model could never have reached. He sees the same pattern ahead for electrical engineering. The work that will disappear first, he notes, is also the work most engineers find least satisfying. “Some of the work nobody’s enjoying doing late evenings before a deadline will probably disappear,” Zein says. “Which is good. I think engineers will not disappear. I think they will, in fact, be upgraded.”

Why Industrial Companies Are Reaching a Structural Limit

The urgency behind AI adoption in electrical engineering is not driven by technology curiosity. It is driven by a convergence of operational pressures that are making the traditional engineering model increasingly difficult to sustain.
Zein maps them out: industrial customers want more customisation per project. Delivery timelines are compressing. Documentation and compliance requirements are growing – and with everything becoming more interconnected (machine software, cybersecurity, regulatory compliance), the overhead on each project increases further. Meanwhile, the engineering talent base is contracting as experienced professionals retire. The result is a structural mismatch: more complexity per project, fewer engineers to handle it, and shrinking time to deliver.
“Many companies are reaching a structural limit with traditional engineering approaches,” Zein says. “That’s why AI is becoming strategically important. Not because it’s fashionable – and yes, it is fashionable – but because the current engineering workflow is under pressure and difficult to scale.”
The 54% finding from WSCAD’s survey is worth returning to here: when more than half of the engineering workforce has no bandwidth for innovation, the problem is not individual productivity. The problem is that the model itself has hit a ceiling. AI, in that context, is not a nice-to-have. It is structural relief.

Reference: How AI is restructuring control cabinet engineering

What System Integrators Should Take From This

Zein gives system integrators a specific message: they are in a structurally strong position, because they sit closer to real operational problems than almost anyone else in the industrial ecosystem.
The challenge for SIs is familiar: margins are under pressure while project complexity continues to rise. AI addresses that squeeze directly – faster engineering throughput, faster documentation, fewer repetitive tasks, better knowledge reuse across projects. As a force multiplier for exactly the kind of engineering-heavy work that defines SI projects, the value proposition is concrete. But Zein makes a distinction that matters: the winners will not be the companies with the most AI tools. “It’s not necessarily the companies with the vast AI tools that are going to win,” he says. “I think the winners will be companies that integrate AI into updated engineering workflows.” The tool is not the advantage. The integration is.
WSCAD’s own involvement with CSIA reflects this view directly. “System integrators are often closest to the actual operational realities inside industrial companies,” Zein explains. “We believe AI adoption in industrial automation will happen through practical implementation – not theoretical discussions.” CSIA, he adds, brings together the organisations that are actively building the future of industrial systems – making it, in his words, “a very valuable ecosystem.”

Reference: How AI is restructuring control cabinet engineering

Why the Real Barrier Is Leadership, Not Software

Three years into WSCAD’s own AI transformation, Zein is candid about what turned out to be harder than the technology. The technical barrier, he says, was smaller than most people expect. “CPUs and GPUs are sometimes less stubborn than organisations.”
WSCAD began building internal AI competence in early 2023, shortly after the initial GPT release in late 2022. What followed was a pattern most organisations will recognise: tension between people eager to move forward and those defending established methods. Zein acknowledges both instincts exist in most people simultaneously – the pull of the comfort zone on one side, genuine curiosity on the other. The deeper challenge was organisational behaviour: “The willingness to experiment, the tolerance for change, the speed of decision-making, and whether leadership actually creates an environment where people can adapt,” as he puts it.
The framing that consistently fails is treating AI as an IT rollout. Organisations that approach it as the next software deployment project miss what the transformation actually demands – rethinking workflows, responsibilities, engineering processes, and sometimes business models themselves. “If you treat AI as an IT topic – ‘it’s the next tool we’re introducing’ – that’s a fundamental misread,” Zein says. “It’s an operational transformation topic.”

Successful AI adopters, in Zein’s observation, share several traits:

• Leadership that actively supports experimentation
• Decision-making that moves faster than the technology itself
• Tolerance for process change, including the discomfort that comes with it
• Willingness to rethink responsibilities and workflows – not just add new tools to old ones
• Organisational learning built through early adoption rather than delayed observation

Zein is equally direct about the remedy: the cultural shift that genuine AI adoption demands takes time to build – and that is exactly why starting early matters. “This cultural shift matters enormously because it takes time,” he says. “Starting early pays off.”

AI Is Not the Advantage – This Is

As AI capabilities become commoditised – available to any organisation willing to pay for access – the question of what actually creates durable competitive advantage becomes more important.
Zein’s answer cuts against the intuitive assumption that having AI is the goal. The real advantage, he argues, is not using AI but how effectively a company can apply it within its specific domain. In electrical engineering, domain expertise is not a soft advantage. It is a hard constraint. The field operates within a dense framework of industry standards, safety regulations, and engineering norms that an AI system must understand and respect to be useful rather than dangerous. Generic AI cannot navigate those constraints without domain-specific training and integration. That gap between having AI and operationalizing it productively in a technically constrained environment is where the actual competitive distance gets created. “The winners won’t simply have AI,” Zein says. “They will understand workflows, customer pain points, industrial requirements, and how to operationalise AI safely and productively. And that takes time – and that’s much harder than just adding a chatbot to your support.”

The implication for companies evaluating their AI posture: the relevant question is not “do we have AI tools?” It is “how deeply have we integrated AI into the specific workflows where our domain expertise creates value?”

Realistic Expectations: Productivity, Not Magic

Zein earns credibility by refusing to oversell. “Companies can expect productivity improvements, but not magic,” he says plainly.
He identifies a pattern that plays out repeatedly across technology adoption cycles: organisations treat AI as a box labelled with all the problems they have not solved, apply the technology, and discover the underlying problems are still there – now running faster. “Bad processes with AI are still bad processes.” The technology does not fix the process; it amplifies whatever is already there.
The sequence that actually works is the opposite of what most organisations attempt: clean up workflows first, structure engineering knowledge, standardise processes – then embrace AI pragmatically within that foundation. Done in that order, the gains are real: faster project delivery, more consistent documentation, lower error rates, and the ability to scale engineering output without proportionally scaling headcount. Zein summarises the potential plainly: “It’s not a replacement for operational excellence. Companies that clean up workflows, structure knowledge, standardise on new processes, and embrace AI pragmatically can absolutely achieve major gains in speed, quality, consistency – and most of all, scalability,” he says. “AI is really a segue to scale.”

Reference: ELECTRIX AI: Artificial intelligence as a productivity game changer

The Compounding Cost of Waiting

The final thread in Zein’s argument is about timing – and it is the most strategically important one for organisations that default to a “let’s wait and see” posture.
The gap between early AI adopters and late movers, five years from now, will not primarily be a technology gap. It will be an organisational learning gap. Organisational learning cannot be acquired retrospectively – it compounds forward, or it does not compound at all. Companies adopting AI early are building internal know-how, operational experience, and cultural adaptability right now. “And that’s not just a fancy word,” Zein says. “That’s actually hugely important.” The engineering team that has spent three years integrating AI into real workflows has built something its competitors cannot simply license.
The most dangerous posture, in Zein’s view, is the one that feels like prudence. “The most dangerous sentence today would be: ‘Let’s wait a few years and see what happens with AI,'” he says. “Because that compounds over time.”

“The biggest difference will be between companies that learned how to evolve while others waited for certainty. And historically, certainty usually arrives too late.”
Asked what advice he would give to a younger version of himself, Zein does not reach for a strategy framework. The answer is the same principle applied personally: “Be bolder. Dare more. Take bets. Don’t listen too much to possible obstacles and worries that others might see in your way – you’ll face them soon enough anyway. I don’t think anyone has ever regretted trying something and failing. Probably most people regret not having tried early enough.”