Three Years Building Our Own China Fulfillment Stack. Here's What's Inside.

Most founders document the journey from day one. The "just quit my job" post. The "we got our first customer" thread. The build-in-public update that gets two thousand likes and zero revenue.

I did the opposite. I went heads-down for three years and built the stack.

That was the right call, and it's why Peregrine ships features in days instead of waiting on a vendor's roadmap. Three years, every layer of the stack built in-house. Here's what's inside, what it took, and why owning the tech is the moat.

It started with a Facebook page and a 16-year-old kid from Macedonia

I made my first dollar online at 16.

Not from a course. Not from a mentor. From a Facebook fan page about cars that I monetized with CPA offers. Dumb? Sure. But it taught me something a business degree never could: the internet does not care where you are from.

I am from Macedonia. A tiny country in the Balkans that most people cannot point to on a map. I had WiFi and time. So I built websites, ran Facebook pages, started dropshipping stores, did affiliate marketing.

By 20 I was making decent money. Nothing crazy. Enough to know ecommerce was where I wanted to live.

The problem: every store I built hit the same wall. The supplier. Late shipments, wrong products, zero communication, quality that looked nothing like the AliExpress photos. Every time I scaled, the backend fell apart.

I kept thinking: there has to be a better way.

Then I went to China and everything changed

In 2023 I was 23. I booked a one-way ticket to Shenzhen.

If you do not know Shenzhen, it is the ecommerce capital of the world. Not Silicon Valley. Not New York. Shenzhen. This is where the people who actually move product live. The ones doing seven and eight figures a month. Not posting about it. Doing it.

Most of them do not speak English.

I got lucky. I made friends with Chinese ecommerce sellers, the guys running massive operations, shipping tens of thousands of orders a day. They took me in. Showed me how things work on the ground. Not the Alibaba version. The real version.

One night changed everything.

The dinner table that humbled me

I am sitting at a dinner table in Shenzhen. Twenty-plus ecommerce sellers. All Chinese. All doing serious volume. Some running factories, some running brands, all moving real product.

Someone asks me how I source. I said: "Alibaba."

The table laughed.

Not in a mean way. In the way you laugh when a mechanic tells you he fixes cars by watching YouTube.

One of them pulled out his phone, opened an Alibaba listing, and pointed at the company name. "See that? It says 'trading' in the name. That is not a factory. That is a reseller."

Then he explained something that completely rewired how I think about sourcing.

That reseller on Alibaba has the same killer instinct as you. He will do everything to sell more. Which means:

• He will offer your winning product to every other buyer he has
• He will publish your product on AliExpress himself
• If he switches the backend factory, because he found a cheaper one or the old one raised prices, you will never know. But your customers will notice.

The quality shifts. The color is slightly off. The packaging changes. Returns spike. Reviews drop. You are sitting at your laptop wondering what happened while your supplier is already selling your exact product to your competitors.

"Alibaba is fine when you are starting," he said. "If you are serious, you need to be here. You need to talk to the factory directly. You need to see the production line."

That dinner taught me more about ecommerce than six years of running stores from a laptop in Macedonia.

The gap nobody was filling

After that dinner I spent months traveling around China. Visiting factories. Sitting in supplier offices. Learning the system from the inside.

I saw the gap clear as day.

On one side: thousands of factories making products for pennies. Good products. Real manufacturers with real capacity. The kind of suppliers my Chinese friends were using to build eight-figure businesses.

On the other side: Western dropshippers and DTC brands paying 3x markups through Alibaba trading companies, getting garbage shipping times, with zero visibility into their supply chain.

The entire system was held together with WhatsApp messages, Excel spreadsheets, and blind trust.

Nobody had built the infrastructure to properly bridge these two worlds in a way that was transparent, tech-driven, and worked at scale.

So I decided to build it myself.

Three years heads-down (the build that became the moat)

Three years in full development mode. A WMS from scratch. A sourcing panel. A user interface. An order management system. A transportation management system. A tracking system. Integration with Chinese B2B APIs. Integration with logistics carriers. Integration with Shopify.

From scratch. Every single piece.

While other fulfillment companies were slapping their logo on Chinese ERP software (dependent on tech they do not control, cannot customize, and cannot scale), the Peregrine stack got built layer by layer. Every line of code. Every database schema. Every API endpoint.

Three years. A dev team in Shenzhen. Testing systems. Breaking things. Fixing them at 3am.

No posts. No build-in-public threads. No "week 47 update." The bet was: build the entire stack first, then ship features fast on top of it. That bet paid off. Peregrine now releases features in days while competitors wait on their ERP vendor for quarterly updates.

The only thing to do differently from here: publish the work. Three years of warehouse calls, supplier wins, dev fires, hard-won infrastructure lessons. From now on, every week.

What we actually built (the tech nobody sees)

Peregrine Ship is not a fulfillment company that uses someone else's software. We ARE the software.

Let me walk you through what sits behind the scenes, layer by layer.

Own WMS (Warehouse Management System)

Every product that enters our Shenzhen warehouse gets scanned, QC'd, photographed, and tracked through our system. Not through some Chinese ERP that crashes when you have more than 500 SKUs. Our system. Built for how we actually operate.

Inbound. Outbound. Inventory counts. Bin locations. Quality control checkpoints. All of it custom.

Own Sourcing Panel

This is the one I am most proud of. We integrated directly with Chinese domestic B2B platforms, the ones Chinese sellers actually use, not the international Alibaba full of trading companies.

When a client's order comes in, our system automatically pulls product data, price, weight, domestic shipping cost, directly from the factory listing.

No more copy-pasting. No more "let me check with the supplier and get back to you in 48 hours." The data flows automatically.

Remember that dinner table in Shenzhen? This is the tech version of what my Chinese friends taught me. Go direct. Cut the reseller. Own the relationship.

Own Tracking System (PST, Peregrine Ship Tracking)

We launched our own tracking network in January 2026. Every shipment gets a PST number, our branded tracking ID that works across the entire journey.

Customers track at track.peregrineship.com. Two modes:

• Full journey: shows everything from China warehouse to doorstep
• Last-mile only: hides the China origin completely

That second one matters more than you think. A lot of DTC brands do not want their customers seeing "Shipped from Shenzhen, China" in tracking. With last-mile mode, the tracking only shows the local carrier, DPD, Yodel, USPS, whatever handles the final delivery. White-label. Branded. Professional.

Where AI actually fits (not where you think)

Everyone is talking about AI in ecommerce. AI product descriptions. AI customer service. AI ad copy.

We use AI for something nobody talks about: customs declarations.

Every package that crosses a border needs a declaration. Product name, value, HS code, material composition. Get it wrong and your shipment gets held at customs. Or worse, gets returned to China. Or worst: your client's customer gets slapped with unexpected duties and leaves a one-star review.

The problem? A single client might have 2000 SKUs. Each one needs a proper English declaration name that customs officers can actually understand. "Blue thing from China" does not cut it. Neither does the Google-translated gibberish that comes off most B2B listings.

We use AI to generate accurate declaration names based on product data. It pulls the Chinese product listing, cross-references the product category, translates it, formats it into proper customs language.

What used to take our team hours now takes minutes. The error rate dropped off a cliff.

Not sexy. Nobody is going to make a TikTok about automating customs declarations. It saves our clients from shipments getting stuck at borders. That is where AI actually matters in supply chain: the boring, critical stuff that breaks everything when it goes wrong.

Why this matters right now (the post-de-minimis world)

Let me tell you what is happening in 2026 that makes all of this relevant.

The de minimis loophole, the one TEMU and SHEIN built their entire business on, got tightened hard in 2025. Section 321 still exists for sub-$800 shipments, but the paperwork and Type 86 filing requirements made the casual abuse impossible. The EU is following with a 3-euro-per-package fee in July 2026.

For lazy dropshippers shipping one-by-one from AliExpress, the margins are gone. That $3 customs processing fee alone kills profitability on anything under $30 AOV.

For smart operators this is the biggest opportunity in ecommerce fulfillment in years.

Here is why: bulk shipping to a China-based fulfillment center, consolidating orders, using proper shipping lines with negotiated rates, saves 35%+ compared to the old direct-ship-every-package model.

The era of "find product on AliExpress, list on Shopify, pray it arrives" is done. What replaces it is infrastructure. Real systems. Real warehouses. Real quality control. Real tracking.

That is exactly what we spent three years building while everyone was posting.

The stack nobody else has

What makes us different from every other "China fulfillment company":

We do not depend on Chinese ERP software. Most fulfillment companies in China, including some of the biggest names you have heard of, run on Chinese ERP systems. Their core tech is owned by someone else. They cannot customize it. They cannot ship features fast. They are tenants, not owners. When that ERP has downtime or decides to raise prices or deprecates a feature, they are stuck.

We own every piece:

• WMS, warehouse operations
• OMS, order management
• TMS, transportation management
• PST, our tracking network
• Sourcing Panel, factory connections
• Client Dashboard, balance, analytics, cost breakdowns, inventory levels
• Shopify App, direct integration, live on the App Store

When a client says "can you add this feature?" we do not submit a ticket to some Chinese software company and wait three months. We build it ourselves. In January alone, we shipped a complete tracking system, new balance logic, new B2B listing API sourcing integration, and a shipping calculator. One release.

That is what three years of heads-down building gets you.

What I would tell 22-year-old me

Three things.

Document everything from day one. The journey is the content. People want to see the warehouse being built, not just the finished building. Three years of warehouse calls, supplier wins, and dev-team fires would have built this brand faster than any ad campaign. Start publishing on day one. Even rough drafts. Even when the system breaks.

The tech is the moat. Every fulfillment company can rent warehouse space and hire packers. That is commodity work. The companies that win long-term build systems. Our tech stack is the single biggest thing that separates us from competitors who look identical on the surface.

Go sit at the table. The best education I ever got was not from a course or a YouTube video. It was from sitting at a dinner table in Shenzhen with people who were 10x ahead of me, and being humble enough to listen when they laughed at my sourcing strategy. Find those tables. Be the dumbest person in the room. Then go build what you learned.

What's next

Every week from here, a breakdown of what we see from inside the machine. Real shipping costs. Real sourcing data. Real warehouse problems. Real client wins. The stuff nobody in this industry wants to show you.

Transparency is the only unfair advantage left in fulfillment.

If you are running an ecommerce brand and shipping from China, or thinking about it, follow along. This is going to be the most honest look at fulfillment you will find anywhere on the internet.

And if you are doing $50K+/mo and your current fulfillment setup is costing you margins and sleep, let's talk. Not a sales pitch. A conversation about what is actually possible when you have real infrastructure behind your brand.