Forbes Technology Council Member and Principal Software Engineer in AI, Harsh Verma, has set out his argument that the AI engineering profession has reached a structural turning point, and what he argues the next generation of engineers must do to remain relevant.
— In a newly published piece on Hackernoon, Harsh Verma sets out why current AI governance frameworks are failing to keep pace with autonomous AI systems, and what organisations must build instead.
As enterprises accelerate adoption of generative AI, a growing debate has emerged inside the technology industry: will the future of AI engineering belong to model builders, or to architects capable of orchestrating intelligent systems at scale?
According to Harsh Verma, Principal Software Engineer in AI at Palo Alto Networks, Forbes Technology Council Member, Stanford Distinguished Scholar, and IEEE Senior Member, has published a framework for the evolution of AI engineering, setting out his argument that the profession has undergone a fundamental structural shift and the industry has already crossed that threshold. His position, set out in a recently published piece on Hackernoon, is that the frameworks currently governing AI systems are built for a world that no longer exists.
The publication, titled “AI Governance Is Failing Because We are Regulating Models Instead of Behavior”, argues that modern AI systems, particularly agentic and autonomous systems capable of reasoning, planning, and acting across enterprise workflows, can no longer be governed through static pre-deployment controls alone. These publications reflect Verma’s position and aspects at the intersection of AI engineering and cybersecurity.
Verma’s central argument in his Hackernoon publication is that AI engineering has moved decisively away from model building as its primary discipline, toward system architecture, integration, and governance. The shift reflects broader industry movement toward agentic systems and enterprise orchestration architectures. In his assessment, the engineers who will define the coming decade are those who can design, orchestrate, and govern intelligent systems across entire organisations, not those who optimise individual models. The shift reflects broader industry movement toward agentic systems and enterprise orchestration architectures
The Shift Harsh Verma Has Identified
Verma says that as large language models, multimodal systems, and agentic AI frameworks have matured, the centre of gravity in AI engineering has moved up the stack. Engineers are, in his framing, no longer primarily responsible for training models from scratch. They are responsible for designing architectures that connect AI across workflows, manage reliability at scale, and build systems capable of evolving as the technology around them changes.
He identifies agentic AI as the development most defining this transition. These autonomous systems, capable of acting, planning, and working toward goals with minimal human intervention, require disciplines that traditional AI engineering training did not anticipate, such as memory management across multiple tasks, tool integration, and reasoning and decision loops.
“AI is no longer limited to building models. The challenge is no longer simply building models, The challenge is designing systems that can evolve, reason, integrate with enterprise workflows, and operate reliably at scale.” Verma mentioned during recent discussions on agentic AI systems. Enterprise AI success increasingly depends less on isolated model accuracy and more on orchestration, governance, and system reliability.
His proposed shift is from regulating models to regulating behaviour, embedding governance into the operational systems that monitor, manage, and constrain AI behaviour in real-world conditions rather than assuming compliance can be established before deployment and maintained without active oversight.
The AI Skills Verma Recommends Engineers Now Develop
Beyond technical skills, Verma’s framework argues that systems thinking, infrastructure awareness, and continuous adaptation have become prerequisites for AI engineers. He also makes the case that communication, personal branding, and the ability to influence stakeholders are now equally important, a position he acknowledges challenges a profession that has historically rewarded technical precision over presence.
“Make sure the tools you are building are comfortable for the users, so they can leverage it in the very best way,” Verma explained, framing engineering excellence as a human-centred discipline as much as a technical one. “The engineers defining the next decade will be those capable of governing and orchestrating AI systems across entire organizations.”
Speaking Engagements, Mentoring and Industry Recognition
His perspective has gained visibility through industry speaking engagements and technical discussions focused on enterprise AI transformation. Alongside his published framework, Verma has been active in sharing these ideas across speaking engagements on topics involving agentic security systems, large-scale machine learning operations, and the operational realities of deploying generative AI in enterprise environments. His talk on the era of agentic security was published on UC Berkeley SkyDeck’s official YouTube channel. He has spoken at TrueML Talks on big data and machine learning practices at Palo Alto Networks, at the Digital Battleground series, and at FutureAGI’s GenAI and Intelligent Agents event. He mentors early-stage ventures through UC Berkeley’s SkyDeck accelerator and advises startups on AI strategy, infrastructure, and implementation challenges.
Recent Recognitions
In March 2026, Verma received a Global Recognition Award evaluating contributions across Innovation, Leadership, Service, and AI and Research. He was confirmed as a Forbes Technology Council Member earlier the same year and placed on the Fortuna Global 100 List in AI. He also got an Outstanding AI engineer recognition by the Noble Technology Awards.
He has also been accepted as an IFGICT Fellow, elevated to IEEE Senior Member status, and awarded the IOASD Royal Fellowship in recognition of his contributions to innovation, leadership, AI, and cybersecurity.
Forthcoming Book
Verma is currently completing AI-Engineering Beyond the Code, a book that extends his published arguments on AI engineering. It makes the case for a holistic approach combining technical depth with leadership, communication, and personal brand. His central argument is that the next wave of AI engineering will belong to those who can govern, orchestrate, and lead intelligent systems as part of a larger ecosystem.
As enterprises continue integrating AI into critical operations, the debate around what defines modern AI engineering is likely to intensify. Harsh’s position reflects a growing view within the industry that the future may belong less to isolated model development and more to the architects capable of managing intelligent systems as part of complex organizational ecosystems.
About Harsh Verma
Harsh Verma is a Principal Software Engineer in AI at Palo Alto Networks, a member of the Forbes Technology Council, a IEEE Senior Member, a Stanford Distinguished Scholar, and a mentor at UC Berkeley SkyDeck. His work focuses on generative AI, large language models (LLMs), AI agents, natural language processing, deep learning, distributed systems, and enterprise-scale machine learning platforms. He is recognized for building scalable AI and cybersecurity systems that bridge research innovation with real-world production impact. Connect with Harsh on LinkedIn, Furthermore, information is available through published talks and industry engagements.
Contact Info:
Name: Gianmarco
Email: Send Email
Organization: Xraised
Website: http://xraised.com
Release ID: 89191464
If you detect any issues, problems, or errors in this press release content, kindly contact [email protected] to notify us (it is important to note that this email is the authorized channel for such matters, sending multiple emails to multiple addresses does not necessarily help expedite your request). We will respond and rectify the situation in the next 8 hours.
