By Jeremy Bunting, VP of Engineering at SERHANT.
History is repeating itself in technology adoption, and most companies don’t realize it yet.
When SaaS applications began replacing on-prem software in the early 2000s, the shift wasn’t just technological, it was organizational. Cloud-based systems came with a completely different lifecycle: continuous updates, configurable workflows, and integrations that could change overnight. That forced the creation of new, specialized roles - Business Systems Analysts, application administrators, integration specialists - people whose full-time job was to keep the technology aligned with the business. I was in some of those rooms, and building the tech to support those goals.
Those professionals had two non-negotiable responsibilities:
- Implementation & Configuration: Translating business needs into system setup, handling data migration, configuring permissions, workflows, and reporting.
- Ongoing Optimization: Ensuring the tools adapted as processes evolved, bridging the gap between IT and business teams, and spotting efficiency gains before anyone else did.
The SaaS era proved one thing: tools don’t run themselves. Human expertise was the difference between a system that delivered leverage and one that quietly decayed.
Fast-forward to today’s AI and automation wave. The technology leap feels just as significant, but the organizational adaptation is lagging badly. Employees - finance analysts, marketers, operations staff - are being told to self-build AI automations, prompts, and workflows alongside their existing jobs.
The problems are already visible:
- AI literacy is uneven: Enthusiasm is high, but structured knowledge of capabilities, limitations, and data risks is low.
- Governance is absent: Without centralized oversight, automations become fragmented, redundant, or outright non-compliant.
- Optimization is no one’s job: Workflows launch and then stagnate, with no measurement or scaling across teams.
The result is predictable: disconnected automations, “shadow AI” processes, and squandered opportunities for enterprise-wide leverage.
The lesson from the SaaS shift is clear - new professional roles must emerge and be embedded into the organization:
- AI Systems Analyst: The modern BSA, mapping business processes to AI capabilities and ensuring alignment with enterprise goals.
- AI Automation Architect: Designing and governing workflows with security, scalability, and resilience in mind.
- Prompt Engineer / Conversational Designer: Crafting and maintaining the language layer for precision and usability.
- AI Ops / AI Product Manager: Overseeing adoption, lifecycle management, and optimization at scale.
AI isn’t removing the need for human implementation and optimization. It’s redefining it. Organizations that formalize these functions now will avoid the chaos of “everyone builds their own thing” and instead run a coordinated, high-ROI AI strategy that compounds over time.
How This Looks in Practice at SERHANT.
At SERHANT., the mission isn’t to “use AI.” It’s to operationalize AI so it becomes a durable advantage for our agents and staff. That means every workflow we automate, every AI-assisted deliverable we ship, and every prompt we refine goes through the same discipline you’d expect from any mission-critical system: governance, measurement, and continuous improvement.
S.MPLE, our AI-powered & advisor-led platform, is a live example of that philosophy. It’s not a chatbot bolted onto the business, while chatbots certainly have their place - especially based on the success we’ve seen with highly personalized versions of them being directly available to our agents and our S.MPLE Advisors. A SERHANT. agent with access to highly personalized AI tools has demonstrated a correlated average increase in GCI of 144% comparing Q2 2024 to Q2 2025, near-identical real estate markets to be able to measure against.
That orchestration doesn’t happen by accident. I work closely with Ryan Serhant, who sets the cultural standard for how technology should enhance - not replace - the human side of this business. I collaborate with CXO Ryan Coyne, whose experiences building and deploying AI and SaaS products that show deep consideration of the people they’re built to serve has shaped S.MPLE’s architecture into something both agent-friendly and enterprise-grade. And I partner with Director of S.MPLE Kelsey Lee, who ensures the operational layer is so seamlessly tailored for the operators behind it that agents can trust it from the first request they make.
As Ryan Serhant often says, “If technology doesn’t make the human experience better, you’re not innovating - you’re just adding noise.” That mindset drives every decision we make in how AI is deployed, maintained, and measured here.
Why I Approach AI Like a Designer
Before I was an engineering leader, I trained as an industrial designer. That background has shaped how I see AI systems not as stacks of code, but as experiences people live with every day. In design, you learn quickly that the wrong friction point - an extra step, a poor fit, an unclear signal - breaks trust. In engineering, it’s the same. The best AI products aren’t the ones with the most features - they’re the ones that disappear into the workflow, anticipate needs, and work the same way every time.
That’s the lens I bring to this moment. AI can be a playground but it’s first and foremost infrastructure. And if SaaS history is any guide, those who treat it that way early will own the advantage long after the hype cycle fades and leaves many companies with AI implementation fatigue and regret.