I spent more time thinking about warehouses and delivery routes than I expected when I started working on supply-chain software.
At Wholesale Mandi and later Origin Fresh, I led product and engineering as co-founder and CTO. We built the systems behind procurement, inventory, fulfilment, and delivery. Wholesale Mandi connected more than 500 suppliers, and at Origin Fresh we brought the iOS, Android, and web products to launch in 100 days.
Those are useful details for a project summary. The part I keep thinking about is how much of the work depended on understanding the decisions people made every day.
An order keeps changing after checkout
With fresh produce, stock, quality, and prices can change before an order is fulfilled. An order in the database is only one part of what the team needs to know.
Someone has to decide what to procure. The warehouse needs to know what is actually available. Delivery planning depends on where those orders need to go and when they can get there. Each of those decisions affects the others.
That makes an apparently small product decision quite involved. Deciding what customers can order means understanding stock and fulfilment zones. Procurement also depends on demand, recurring orders, and what suppliers can provide. You have to follow an order through the business to understand what the product needs to account for.
Working through these questions made me pay closer attention to where information came from and who depended on it next. A stock figure can look perfectly clear on a screen while leaving the person making a decision with more questions.
Spend time with the people using it
I worked with procurement and delivery teams to understand their daily decisions, including what happened when orders didn’t go to plan. That helped me understand what our software needed to make visible and where people needed room to make a judgment.
I’ve become cautious about deciding how a workflow should work before understanding the existing one. It’s easy to draw a neat sequence of steps. It takes more attention to find out why someone does them in a different order, or what information they’re waiting for.
For me, that’s also where backend engineering gets interesting. You have to work out which information needs to be current, how a change should reach the people affected, and what happens when part of the process is delayed. The answers come from the operation you’re building for.
I left Origin Fresh in December 2025. Supply chains are still one of the areas I’m drawn to, partly because there’s so much useful engineering work tucked into those everyday decisions.