Aamir Zargham (Lightspeed): Rebuilding long-range planning in 60 days

Lightspeed is the commerce platform that keeps retail and hospitality merchants running, from the checkout counter to the stockroom.

For Aamir Zargham, the company's Director of Strategic Finance and Insights, it all comes down to one recurring question: where should the next dollar go? At a public company chasing growth and profitability at once across multiple portfolios and countries, almost every decision turns into a question about capital.

Aamir had built the company's long-range planning model himself, and for years it lived in a single Excel file that only he could run. To change anything, someone had to message whoever else had it open and wait their turn to get in.

Rebuilding a two-year-old scenario planning process in 60 days

By the time Aamir replaced the Excel file, it had been rebuilt five times and had reached version 1,000, with dependencies running both upstream and downstream. A model juggling 500 or 600 intersections had to stay light enough that every input landed in under five seconds, so speed and accuracy were always in competition.

When a leader asked something the model couldn’t handle, the answer was to come back next week, by which point the decision had been made without finance in the room. And if Aamir was out at quarter-end, the file went stale.

So he rebuilt the entire long-range planning process in Pigment in just 60 days.

Aamir knew the existing model like the back of his hand, so he already knew exactly how the new model should be built. But speed also came from the tools themselves. Aamir thinks about financial modeling the way some people think about Legos, and Pigment's modular UX landed for him right away. Once he had the syntax down, the team moved from nothing to roughly 700 components over those two short months.

As the cost of curiosity dropped, the whole business leaned in

The rebuild changed what scenario planning meant for Lightspeed. Where the old model stretched to handle five or 10 scenarios, the new one let Aamir run more than 80. He sees this abundance of scenarios as scaffolding for the three he carries into a leadership meeting. The real work is narrowing everything down to the strongest signal, building the narrative, and anticipating which questions the C-suite will ask before they ask them.

Aamir’s speed has reshaped how company leaders engage with finance. Spinning up a new scenario used to cost enough that leaders asked the safe, “need-to-know” basic questions, because anything more would take days. Now Aamir opens the model in a meeting, changes a few inputs, and moves the numbers in real time in front of everyone. A leader can ask what happens to the P&L if a particular move reduces churn by some amount at some cost, and the answer appears live in the room.

Getting AI-ready before AI-excited

When it comes to AI, Aamir is patient on the principle that bad data in means bad data out. The work is to build a systems-first pipeline that surfaces the strongest signal, with AI sitting at the tail end to magnify a clean input.

He’s experienced a couple of awe-inspiring moments with AI, and both came from inside Pigment. The first was Pigment’s Analyst Agent, which he watched run end-to-end on a schedule, with an agent and a supervisor coordinating between themselves. The second was Pigment’s Modeler Agent, which the team now uses to update the metadata, comments, and descriptions inside the model so future changes come more easily. Aamir draws a line between automation and agents. Automation repeats the same mechanical steps every time. An agent has room to make decisions across the workflow you assign it.

That distinction shapes the way he delegates. The agent takes on the busy work and surfaces its version of the strongest signal, or the outlier worth chasing across two metrics. A human picks up from there to make sense of the why, applying their seasoned business knowledge and the contextual reality of the industry.

For anyone just starting out, his advice is to hold off on committing to a single tool this early and commit instead to how you plan around it. If the data is too messy for your own people to work with, AI will not be able to fix it.

In financial planning, sharing knowledge can elevate an entire team

If transformation taught Aamir one thing, he says it starts with himself.

Aamir had taken pride in being one of the few people who knew the model from end to end. But letting go of that role and handing the model to others elevated the whole team. The single point of failure is gone, owned now by three people rather than one. Aamir is free to focus on strategy and how finance communicates with the business. And his team has risen to the occasion.

Aamir sees the same pattern shaping the next few years of finance leadership. He believes that the leaders who thrive will think in systems and treat people, processes, platforms, and partnerships as one connected whole. He also believes they’ll build cultures that empower people from many different backgrounds and communicate with a clarity that respects the reader’s time.

When Aamir hires today, the quality he looks for first is fluency in both finance and data. Teams are often starved for decision-ready data, and finance sits at both the start and the end of most processes, where any defect gets magnified into a last-minute fire drill. The people who can trace where data comes from and turn it into insight, then into impact, are the ones who move the business at scale.