A conversation with Sabareesh Kappagantu
In consumer software, UI and UX are often about speed, delight, and engagement. In industrial software, they carry a very different weight. When systems support utilities, energy, or manufacturing operations, design decisions directly affect safety, reliability, and real-world outcomes. Sabareesh Kappagantu, a Software Engineering Manager working across AI and large-scale industrial platforms, approaches UI and UX with that responsibility at the forefront.
For Sabareesh, the difference starts with how interfaces are framed. Working on software that impacts utilities, energy, and manufacturing fundamentally changes how he thinks about UI and UX. In these environments, the interface isn’t just a layer of polish on top of functionality; it’s part of the system’s safety, reliability, and decision-making surface. Operators are often responding to time-sensitive situations with incomplete information, and the cost of misunderstanding, delay, or misinterpretation can be high. That reality forces a different standard: clarity over cleverness, predictability over novelty, and trust over engagement.
Unlike consumer software, where UX often optimizes for speed, delight, or habit formation, industrial UX is about reducing cognitive load under pressure. Interfaces need to surface the right information at the right time, make the system state unambiguous, and support confident action rather than exploration. Consistency matters deeply, not just visually, but semantically so that users build reliable mental models over years, not sessions. Good UX in this context often looks invisible because it avoids surprises and minimizes the number of decisions a user has to make when it matters most.
That philosophy becomes even more important when platforms surface massive volumes of real-time data. In Sabareesh’s experience, the challenge isn’t access to information; it’s deciding what deserves attention. Operators don’t need more data in the moment; they need the right signals to maintain situational awareness and act with confidence. That means designing interfaces around intent and risk, not around everything the system can emit.
In practice, this comes down to separating state, change, and anomaly. Operators need a clear, stable view of system state to stay oriented, but they should only be interrupted when something meaningfully deviates from expectations. Trends, raw metrics, and historical detail still matter, but they belong in layers that can be pulled in deliberately rather than pushed constantly. This helps avoid alert fatigue and prevents users from overreacting to noise while still preserving depth when investigation is required.
Time and context also play a major role. What needs to be visible during steady-state operation is very different from what’s needed during an incident or recovery. Good industrial UX allows information to move between foreground and background as conditions change, without forcing users to hunt for it. The goal is to respect the operator’s attention, surfacing information that supports immediate decisions while keeping the rest accessible, consistent, and trustworthy when deeper analysis is needed.
A clear example of translating complexity into usable UX comes from Sabareesh’s work on the asset status capability within the AVEVA Data Hub. Industrial systems generate continuous streams of telemetry, but operators don’t think in raw data. They think in operating modes. Whether an asset is running, stopped, in maintenance, or degrading, and how it transitioned between those states, matters far more than individual data points.
To address this, status was elevated to a first-class concept in the interface. Underlying data and rules were mapped into meaningful operating modes that could be represented visually. Color-coded legends allowed operators to immediately identify warning or critical conditions without interpreting numbers. A time-based Status view then showed how assets moved between modes, making patterns like prolonged degradation or intermittent failures visible at a glance.
That same UX logic scaled to fleet-level views. Status was integrated directly into asset lists, allowing operators to filter large sets down to only those in problematic states. Instead of scanning dashboards or building custom queries, the interface reorganized itself around urgency. Shareable, dynamically updating filtered views reduced friction further by allowing context to be passed to maintenance or engineering teams without forcing them to navigate the entire system.
Onboarding is another area where industrial UX demands a different approach. The complexity of these systems is real and unavoidable. Sabareesh doesn’t believe in hiding it. Instead, onboarding should help users build accurate mental models quickly. One effective pattern is designing around conceptual landmarks rather than features, anchoring users on core ideas like asset, status, time, or state, and letting everything else layer naturally on top.
Progressive disclosure plays a role, but not in the consumer sense of concealing functionality. Instead, depth is exposed through intent. New users should be able to answer basic questions without configuration or deep navigation, while still having clear paths into advanced views when they are ready. Consistent terminology, stable visual encoding, and repeatable interaction models help users transfer understanding across the system without relearning concepts.
As AI becomes more embedded in industrial platforms, Sabareesh is cautious about how intelligence appears in the UI. In high-stakes environments, AI should feel assistive, not opinionated. He prioritizes intelligent features that clarify or contextualize what the system already knows, summarizing patterns, highlighting deviations, or explaining the state, rather than immediately offering recommendations or automation.
Avoiding overload comes down to respecting attention. AI should activate in response to user intent or rising uncertainty, not compete constantly for attention. When it does surface insights, it should be obvious why it appeared and what data it’s based on. Predictability and explainability matter far more than impressiveness. Trust is earned when AI makes ambiguity explicit rather than hiding it.
Across all of this work, Sabareesh treats UI and UX as inseparable from system behavior. Interfaces are not just how users interact with software; they are how users understand and trust it. In industrial environments, good design isn’t about standing out. It’s about holding up under pressure, supporting clear thinking, and staying reliable when it matters most.
