Gandhi Kumar, Senior Staff Incident Commander at Aurora Innovation, has been recognized across major industry publications for his ongoing contributions to incident management, on-call culture, and responsible AI integration in technical operations. Named one of the Top 50 People Making the World More Reliable by Rootly, Gandhi continues to shape how modern engineering organizations handle complex, high-consequence system failures.
In recent publications for the Disaster Recovery Journal, Humans of Reliability, and Uptime Labs, Gandhi outlines actionable frameworks for reducing responder friction and managing high-stress outages. His writing highlights the critical importance of the "Golden Window," the initial 15 minutes of an incident, and defines clear boundaries for artificial intelligence in SRE, demonstrating how automated tools should accelerate context discovery without taking unverified actions during live disruptions.
Artificial intelligence has become one of the most prominent technologies in today’s business environment, enabling quick decision-making, efficiency, and the processing of large amounts of data that no professional could handle alone. As companies use artificial intelligence more widely to support decision-making, one question remains: when must human expertise be irreplaceable?
For the technology professional Gandhi Mathi Nathan Kumar, that question sits at the center of how AI should evolve within incident management. While automation has transformed how organizations identify and respond to operational disruptions, he believes the conversation should extend beyond technological capability and focus equally on the people who interpret information, understand context, and make decisions when circumstances become unpredictable.
"Artificial intelligence is incredibly powerful at identifying patterns and processing information, but critical incidents are rarely defined by data alone. They require judgment, context, and an understanding of human consequences," says Gandhi Mathi Nathan Kumar. "The future isn't about choosing between people and AI. It's about building systems where each strengthens the other."
That perspective reflects an increasingly practical discussion taking place across industries. Organizations are investing heavily in AI-powered monitoring and incident detection, yet operational leaders still recognize that technology performs best when experienced professionals stay actively involved throughout the decision-making process.
Modern incident management depends on rapid access to accurate information. AI can identify anomalies, surface correlations, and reduce the time needed to detect emerging issues. These capabilities allow organizations to respond more quickly than ever before.
However, speed alone does not guarantee the right decision.
Critical incidents often involve incomplete information, changing circumstances, competing priorities, and consequences that extend beyond technical performance. Human judgment becomes especially valuable when organizations must balance operational continuity, customer impact, regulatory requirements, and ethical responsibility at the same time.
For Gandhi Mathi Nathan Kumar, AI should be viewed as an enhancement to professional expertise rather than a replacement.
"The value of AI isn't that it removes people from the process," he says. "Its value comes from helping experienced professionals make better-informed decisions when every minute matters."
Rather than replacing operational leaders, AI enables them to spend less time gathering information and more time evaluating options, communicating effectively, and managing outcomes that technology alone cannot fully understand.
Discussions about artificial intelligence often become polarized, framing technology either as an inevitable replacement for human work or as something to approach cautiously at every stage.
Gandhi Mathi Nathan Kumar believes reality is considerably more nuanced.
His professional experience has reinforced the importance of combining technological advancement with practical operational knowledge. Rather than treating AI as a destination, he sees it as another tool to support experienced professionals who understand both technical systems and the broader organizational environment in which they operate.
That balanced perspective has become increasingly relevant as organizations seek practical guidance rather than theoretical debate.
"Technology should reduce complexity, not create new uncertainty," Gandhi Mathi Nathan Kumar explains. "The best AI implementations are the ones that fit naturally into existing decision-making processes while giving professionals better visibility into what's happening."
This emphasis on collaboration reflects a growing recognition that successful AI adoption depends not only on technical capability but also on organizational trust. Employees are more likely to embrace AI when they understand how it supports their expertise rather than diminishes it.
As AI technologies improve, discussions about responsible use become as critical as the innovation itself. Today, companies realize the need for governance, transparency, and a well-defined relationship between automated and human decision-making. Although algorithms can make decisions based on available data, people accountable for the company's performance will remain responsible for the results.
For Gandhi Mathi Nathan Kumar, maintaining that balance will become increasingly important as AI systems become more sophisticated.
"Every new technological capability should strengthen human decisions rather than replace it," he says. "Organizations achieve better long-term outcomes when technology and experienced professionals work together instead of competing for responsibility."
This shows how such a philosophy can apply to one of the major technological shifts companies are going through. Rather than wondering whether AI is going to replace human intelligence, Gandhi Mathi Nathan Kumar recommends that companies consider how to design systems where computational power works together with professional wisdom.
For real progress in incident management and decision-making, the best results will not necessarily come from choosing between humans and technology. The best way forward is to let each do what it does best.