At 16, I Fired 13 People. Not Because I Wanted To.

The Business That Ran on Paper and People

In 2014, my family's transport business in Kathmandu employed 17 people across operations. Four accountants processed cargo bills by hand — every waybill typed on a desktop, cross-referenced against a ledger, filed in a cabinet. Nine people managed client outreach: cold calls, relationship management, following up on overdue invoices by physically visiting clients. Two more handled dispatch scheduling on a whiteboard.

The business worked. It had worked this way for 20 years. Nobody questioned it because nobody had reason to — until cargo volumes doubled and the manual processes started cracking under the load.

I was 16. I had just discovered Excel.

What I Saw That Nobody Else Was Looking For

To me, the business looked like a data problem. Four accountants spending 6-8 hours each on billing that was fundamentally repetitive — the same fields, the same calculations, the same format, every time. I asked one of them to walk me through the process. He did. I timed him. A single cargo bill took 23 minutes from client order to filed document.

I built a template in Excel that took the same information and produced the same output in under 2 minutes. Then I connected it to a simple database so every bill was logged automatically. Then I automated the invoice follow-up with a scheduled email system (at the time, this felt like magic).

The billing process that took 4 people dropped to 1 person with my tools. The client follow-up system that required 9 people became a semi-automated workflow managed by 2. The dispatch board became a digital schedule anyone could update.

Over 18 months, 13 roles became unnecessary. Some people retired. Some moved to other departments. Some left. I didn't fire anyone — my father did, because the business couldn't sustain staff whose work no longer existed.

I didn't feel powerful. I felt responsible in a way I wasn't ready for. Automation is easy. What it does to people is complicated.

What the Business Looked Like After

Revenue per employee increased by 340% over 3 years. Invoice processing errors dropped to near zero. The business expanded routes from 3 to 9 corridors without hiring a single new back-office role. My father still talks about it as the moment the business survived instead of collapsing under its own weight.

What struck me then — and what I've carried into every role since — is that the automation didn't hurt the business. It hurt the transition. The system itself was strictly better. The challenge was human, not technical.

Nepal to Sydney: Following the Data

I came to Sydney in 2022 to do my MSc in Data Science at Macquarie University. The scholarship was validation that the direction was right. The coursework confirmed what I'd already suspected: the tools change, the problem doesn't. Data is always the same problem — too much of it, in the wrong format, being read by people who need answers not numbers.

I graduated with an 8/10 GPA. More importantly, I graduated knowing exactly what kind of work I wanted to do: the kind that makes the spreadsheet obsolete. Build a system good enough that nobody needs to do by hand what a machine can do better.

What Those 13 Roles Taught Me About Every Engagement Since

Every automation project I take on now starts with the same question: who is doing what, and why are they doing it that way? Not because I want to eliminate people — the opposite. Manual work is expensive, error-prone, and soul-crushing. Freeing people from it is a kindness, if you manage the transition well.

The projects I'm proudest of — the ETL pipelines, the Power BI models, the Azure Fabric implementations — all started the same way my first Excel billing template did. Someone was doing something by hand that shouldn't require a hand. I found a way to close that gap.

• Every repetitive task is a dataset waiting to be automated
• The cost of bad data is always higher than the cost of fixing it
• The best dashboards answer questions people didn't know they had
• Automation that doesn't consider the human transition isn't done

The Instinct That Never Changed

At 16, I saw a cargo bill taking 23 minutes and thought: it shouldn't. That instinct — that friction is a bug not a feature, that manual labor on a repeatable task is a system failure — has never left me. It just found bigger problems to solve.