Businesses adopting artificial intelligence (AI) should start with a measurable goal and judge success by the improvement it delivers rather than the number of AI agents deployed, as the technology moves from helping employees to doing work for them, a senior Oracle executive has stated.
Natalia Rachelson, Senior Vice President of Applications Development at Oracle, noted that the number of agents deployed matters less than the operational improvement they achieve. Business leaders are advised to pick a specific, measurable outcome—such as reducing the time spent resolving accounting errors or lowering transport costs—and establish a baseline first to assess whether AI is making a meaningful difference.
According to Rachelson, companies must get the operational basics right before scaling AI. This includes ensuring reliable data, establishing clear ownership of processes, setting appropriate controls, and training employees to work alongside AI agents. The greatest gains typically occur in parts of a process where work frequently experiences delays or requires repeated manual intervention.
Addressing common misconceptions, Rachelson highlighted that the biggest error business leaders make is assuming AI is merely a tool to do the same work faster, better, and cheaper. Instead, she described the technology as a way to redefine how work gets done. She illustrated this by noting that a workflow involving 10,000 steps might be reduced to just five steps if AI handles everything in between, fundamentally changing the business process.
Earlier AI tools focused primarily on answering questions, drafting content, and making recommendations. In contrast, newer AI agents are capable of coordinating next steps and carrying out authorized actions within a business process. For example, a company aiming to increase its customer renewal rate by two percent could set this as a goal, prompting a team of specialized AI agents to analyze data, formulate a plan, and execute it.
In hospital systems, AI can assist by scheduling nurses with different specializations and shifts, automatically redrawing schedules and alerting replacement staff if a nurse calls in sick. The quickest results are expected in routine processes where teams spend significant time gathering information and chasing follow-ups. In finance, AI agents can reconcile accounts prior to closing books; in human resources, they can flag potential payroll issues; and in supply chains, they can compare shipping options or allocate scarce materials while human teams review and approve the plans.
Rachelson emphasized that people remain in charge throughout these operations. Enterprises must define what an agent is permitted to do, how its output will be checked, and when human intervention or approval is required.
Having been with Oracle for about 15 years, Rachelson compared the current AI shift to previous technological transitions, noting that while AI represents a bigger shift than the move to cloud computing, the underlying questions it raises are familiar. She drew parallels to the anxieties surrounding Y2K at the turn of the millennium, observing that major technological transitions frequently generate extensive questions and concerns before standardizing into everyday business operations.
"The shift toward outcome-based artificial intelligence adoption is a critical evolution for enterprises and growing businesses. As Natalia Rachelson points out, deploying AI merely to accelerate existing manual processes misses the core value of the technology. Business leaders must focus on process redesign, establishing clear data baselines, and setting rigorous governance frameworks. True ROI in enterprise AI comes from solving operational bottlenecks and achieving targeted, measurable business outcomes rather than accumulating deployed tools." — Dr. Shishir Gupta, Founder & CEO, StartupLanes
Recent StartupLanes Articles
Browse through our 30 latest publications on venture capital, startups, and angel investing.