As AI dismantles the traditional data career ladder, the journey of a senior data scientist offers a blueprint for what’s next.
When Ziyu Wang first entered the data analyst profession, the career path was readable: data engineers built pipelines; Data Scientists built models; Product managers defined what to build and why. Roles were distinct, transfers were formalized, and success meant moving up the ranks within your lane.
Wang did not stay in his lane.
During his career at leading organizations in the technology and financial services industries, Wang has served simultaneously as a senior data scientist, engineer, and product manager – building production-grade data infrastructure and advanced analytical models, developing deep cybersecurity expertise, and delivering security tools with significant organizational impact. He designed internal training programs, was invited to lecture at top universities, and became a strong advocate for rethinking how the industry develops talent.
Today, as artificial intelligence erases the boundaries between data sciencein engineering and product management, the career Wang has built crossing those boundaries feels less like an outlier and more like an insight.
The plan, part one: going deep
AI-powered tools have enabled product managers to write their own queries, engineers to create data pipelines with co-pilot tools, and data scientists to deploy models without waiting for technical support. Most businesses reacted to this change by investing in AI tools. Wang says they are neglecting an equally crucial investment.
“Almost no one invests in specialized training in the same way,” Wang says. A large language model can write a query in seconds, but if the the person asking does not understand the profession logic or the way a metric is defined, the result will be technically impeccable and fundamentally flawed. Trash in, trash out.
That’s why Wang’s first principle for navigating the transition to AI is counterintuitive in an era that celebrates generalists: go further.
“Mastering AI is table stakes,” says Wang. “What AI can’t replicate is true domain expertise – the contextual understanding that tells you what question to ask, not just how to answer it. »
Wang’s own career confirms this point. As a data scientist working behind code vulnerability detection tools and a certified ethical hacker, he has developed a deep understanding of how security vulnerabilities manifest themselves in the code — the models that make certain vulnerabilities serious, the conditions under which they become exploitable and the gaps that existing detection methods overlook. This domain expertise was instrumental in designing improvements that more effectively surfaced critical risks, allowing them to be addressed before vulnerabilities could be exploited.
“A general purpose tool data scientist We could have worked on the same tools and provided competent analysis,” Wang reflects. “But without understanding the security domain (how attackers think, what vulnerabilities are actually important, what the tools are trying to detect), you optimize metrics without understanding the impact. Domain knowledge told us where to look.
The plan, part two: going further
The depth of the domain is Wang’s gap. The extent of his abilities is his leverage.
“Going deep is what AI can’t replace,” he says. “Going offshore is what AI now allows you to do.”
According to Wang, data professionals should stop see themselves as service providers to product teams and start building products themselves. They already understand the backend data architecture, analytical frameworks, and business logic. Thanks to AI-assisted development, they can act directly on this knowledge.
Wang’s own work on vulnerability detection tools illustrates this principle. Rather than just the data science layer – providing models and reporting results – he operated throughout the product lifecycle: designing the analytical methodology, engineering the underlying infrastructure, and defining the roadmap that shaped how tools surfaced and prioritized risks. He led end-to-end work that typically involved separate data science, engineering, and product teams.
“If I had just done data science – done the analysis and turned in a slideshow – it would have been added to someone’s backlog,” Wang says. “Instead, I helped shape the product. That’s what ‘going big’ means in practice.”
Preparing the next generation
Wang’s influence on the profession extends beyond the organizations in which he worked. As a recurring guest lecturer at the University of Pennsylvania, the University of Cincinnati, and Wake Forest University, he advises students on building careers in a field being redefined in real time.
Do not optimize for labor market that existed when you started your program,” Wang tells them. “Learn the tools, but spend as much time understanding a field that interests you. The tools will change. Domain knowledge compounds.”
This is advice based on a specific vision of where the profession is heading. Within three years, Wang predicts, the traditional separation between “data team” and “product team” will be meaningless in most technologies businesses. Professionals who have embraced both depth and breadth will occupy hybrid roles: parts analyst, parts engineer, parts product builder. Those who resisted will see their role automated or absorbed.
It’s a future Wang has been building toward for years — not by predicting it, but by living it.
“This is the biggest opportunity the data profession has ever had,” Wang says. “For the first time, one person with domain expertise and AI-powered tools can do what previously required a team of five. whether you grab this – or keep it while waiting for a ticket in the backlog.
