Ashwin Raj, CEO of LeafLink, joined us for an interview on the need for businesses to transition from static systems to intelligent, autonomous workflows.
Raj highlighted the “insight-to-action gap” as a major challenge, advocating for impartial systems to bridge this gap.
He advised a cautious approach to AI integration, starting with human oversight and gradually increasing autonomy.
Raj envisions AI elevating human roles, focusing on high-value tasks rather than replacing jobs.
The era of static enterprise data is over. For decades, businesses have relied on rigid, linear systems that require a slow, manual process of human intervention. The result is a bottleneck of speed and innovation.
We spoke with Ashwin Raj, a seasoned tech executive with experience scaling multi-billion-dollar platforms as the Head of Rideshare at Lyft, as well as tenure at ezCater, Amazon, Visa, and Booz Allen Hamilton. Raj is currently the CEO of the software ecosystem for wholesale cannabis commerce LeafLink, and brings a well-rounded perspective on how to manage the enterprise transition in any climate. He offered a guide for leaders to turn brittle, one-dimensional systems into intelligent, autonomous workflows.
Beyond static systems: “Humans are pulled in a thousand directions. We have so much going on, but an agent focused on individual elements, and agents collaborating with each other brings about a whole new level of productivity and efficiency and speed that is unmatched today by any static system,” Raj said.
The human constraint: According to Raj, the challenge facing modern businesses isn’t a lack of data. The problem lies in the “insight-to-action gap” created by human-centric challenges. “The constraint is often the individuals who are tasked with analyzing it,” Raj stated. “We are subject to our own biases and preferences, and sometimes people don’t want to present bad news and things that are going wrong.”
Impartial agentic systems: This isn’t a moral failing; it’s a human reality. The solution, he proposes, is to empower an “impartial system” to bridge this gap. However, he adds a critical caveat: the system is only as good as the training data it was built on. The goal is to train agents on quality data and then empower them to make objective recommendations and take action.
For leaders navigating this transformation, Raj offers a practical, step-by-step playbook grounded in caution and control. It begins with a simple principle: start slow.
The early days: “You don’t jump all in on day one; you build your way into it,” he advised. “In the early days, you need to have a higher degree of human oversight to make sure that these agents are making the right decisions.”
Human-in-the-loop guardrails: This “human-in-the-loop” learning process is the core mechanism for the rollout. When an agent encounters a situation it hasn’t been trained for, a human provides the input. Over time, as the agent experiences more use cases, it becomes more autonomous, and the need for constant oversight diminishes. To manage the inherent risks, Raj champions the use of a powerful framework.
AI and data governance: Proper governance frameworks are especially critical when agents interact with external data systems, particularly those that share private or medical data. “You need to put guardrails in place that dictate the actions these systems can take, and what actions they should not take.”
This shift naturally raises the question of what happens to human jobs when the still-nascent tech progresses even further. Raj’s answer reframes the entire human-AI relationship from one of replacement to one of elevation.
Robotic members on a human team: “Our roles change to become those managers and guides of these agents,” he said, offering a powerful analogy. “It’s like a manager and your team. Think of agents as the members of your team, and you’re guiding them.” This new role involves specific, high-value responsibilities: “ensuring that they’re making the right decisions, the right choices, checking on it, avoiding risk.”
Up-skilling: He uses the example of a finance department: while a company may no longer need 100 auditors for repetitive tasks, it will still need skilled auditors to handle new situations, work with customers, and train the agents. The conclusion is optimistic. “Our role gets elevated to doing more value additive work rather than all the mundane repetitive tasks that we have to do.”
For C-suite leaders facing pressure from the board for speed and from their teams for clarity, Raj offers a core strategic principle to cut through the noise. “I would call it prioritization more than sequencing,” he explained. He explained that every AI initiative must be strictly aligned with the company’s core strategic priorities, whether those are “efficiency outcomes” or “new product outcomes.” This provides a logical, defensible rationale for every decision, satisfying both the board and the team. It also empowers employees, who “appreciate it and they know.
Starting from step one: To maintain confidence, leaders must also communicate a multi-step vision. “As a leader, you have to reassure the board, ‘This is step one, but here’s how I’m planning to do steps two and three using AI capabilities as well.'”
Looking ahead, Raj sees the impact of this transformation extending far beyond the boardroom, fundamentally changing day-to-day life. He offered a tangible example-turned-analogy from his own field to illustrate how AI elevates human creativity rather than replacing it. “I’m really inspired by the future is the cloud kitchen concept with automated capabilities to build Michelin-class quality food,” he envisioned. “Now, what happens to the Michelin chefs? They’re going to go one step further in creating more types of food that challenge and strike your palate in a different way.”
This principle, he believes, will scale to solve humanity’s greatest challenges. His grand vision is one where technology pushes us all to greater heights. He optimistically concluded, “What are we if we are not optimistic about the future of the world?”
