Supporting Analysts Means Clearing Out Rote Tasks to Support Problem-Solving and Storytelling

There was a time when knowing SQL and Python would get you the interview. In today's market, however, hiring managers look for how you think about the business. That same instinct separates analysts at every stage of the job, from navigating infrastructure and evaluating AI tools to building governance habits and telling stories that actually land with leadership. Sohan Sethi knows the difference when he sees it.

As Unit Manager of Analytics and Reporting at Health Care Service Corporation (HCSC), he leads a team auditing 1.2 million healthcare providers in a HIPAA- and CMS-regulated environment. Before moving into the enterprise space, he co-founded two startups by age 20, giving him a front-row seat to how analysts add value at different stages of a company's lifecycle. With a 150,000-plus LinkedIn community and experience serving Fortune 500 clients, he spends a large share of his time hiring and developing analysts, watching which skills actually move careers forward.

What are those skills? For Sethi, they understand that analytics work in a larger business context: "The analyst who is able to bridge the gap between a business problem and the technical skills is the one you want to be working on your team.”

• The analyst gap: When it comes to how average analysts approach a new problem, Sethi says the pattern is predictable. "A medium or average-level analyst will help you answer the question you asked them. A high-level analyst, however, will help you with questions you should ask. Or, at least, help you shouldn't be asking the questions you're asking."

When analysts work with millions of customers and their data, the systems are just as important as the code. Unfortunately, many teams encounter friction during the transition from legacy systems to a unified data architecture, even when the end goal is sound. Sethi says the underlying storage and compute environment affects the workflow, noting that using next-gen data infrastructure can affect analysis speed. Analysts are not always expected to become infrastructure experts, but they often benefit from understanding the limits and strengths of the systems and storage architecture they rely on.

• Terabytes and tunnel vision: Sethi says his own thinking on infrastructure awareness has shifted meaningfully in recent years. Where he once would dismiss architecture as a concern for analysts, he now sees it as an essential part of his work. "A few years ago, my answer would have been that I shouldn't be looking at storage; it doesn't really matter. Over the past two to three years, my answer has definitely changed. Outdated systems can actually limit the ability of the analyst and the team to perform the analysis to the fullest extent."

• Managers as a buffer: That awareness extends to access and security, which Sethi frames as a leadership responsibility rather than an individual burden. "It is the leader's job to make sure the right constraints are set in place, but at the same time, the right access is given to those analysts."

That pragmatic approach extends to artificial intelligence. As organizations invest in improving the state of AI and exploring new ways of transforming enterprise data with AI, data teams often feel pressure from top-down mandates. Sethi is seeing a massive push toward highly custom, in-built applications within organizations. He advocates moving past AI hype toward evaluation, focusing on concrete bottlenecks regardless of how executives are thinking about market trends.

• Dodging the bandwagon: Sethi says the pressure to adopt AI can cloud judgment about where it actually adds value. "To put it bluntly, people are hopping on the bandwagon with the mindset that everyone must use AI or risk being left behind." As a manager, he often breaks broad initiatives into targeted pilots for his team to "limit going 10 steps backward" and instead "realign one small thing into the system" before expanding.

• Delegating the drudgery: Where AI does earn its place, Sethi says, is in eliminating the repetitive work that keeps analysts from higher-order thinking. "Redundant tasks like pulling data from a system or performing data cleaning actions can be offloaded to an AI system. It just gives you so much more time to think about the larger business problem."

Sethi observes a similar pattern in how organizations approach governance. Drawing on his startup background, he notes that some early-stage teams operate with zero governance policies, using AI tools freely with customer data and intellectual property. On the other end of the spectrum, many mature enterprises struggle to avoid treating compliance as a rigid checklist. Effective automated data governance and human accountability often rely on simple, explicit guardrails that people can actually follow, especially at a time when research suggests many companies are adopting AI faster than they are governing it.

• Moving past performative: In his own work, Sethi advocates for responsible AI governance rooted in planning for the next five years. "I'm hearing from some folks that there's a lot of performative governance. And I ask, how can an analyst move from just checking a box or following a checklist to actually using governance as a tool to make decisions faster?"

As AI takes on more baseline data extraction and governance protocols tighten around sensitive information, the analyst’s day-to-day work looks different. Tools change constantly, but Sethi observes that one of the most reliable ways to stand out is the ability to deliver a compelling narrative that connects the numbers to the business.

• Pretty presentation syndrome: Sethi says the bar for what constitutes "good enough" analysis is deceptively low. What separates fast-rising analysts from the pack is storytelling. "Any analyst can come up with data. Any analyst can just pull data from SQL Server, put it in a pretty-looking presentation, and submit it to leadership. But the person who is able to present the findings with a story such that it connects to the larger business outcome is actually helping the team."

From hiring to governance to how findings get presented, Sethi keeps coming back to the same quality: an analyst who starts with the problem, not the data. As tools accelerate and organizations layer on new constraints, that instinct to ask why before asking how becomes harder to teach and easier to spot. Sethi sees it as the quality that compounds over time. "The delivery of contextual insights and storytelling skills really enhances the output of the analysis," he says. "Just looking at what we're really trying to do rather than just answering some question that was given to me, that's what actually helps the larger business."