E-waste isn’t just old electronics headed for the landfill. It’s a gold mine — literally.
End-of-life computers, smartphones, and circuit boards contain precious metals like gold, silver, copper, and palladium. Extract these materials properly and you protect the environment from hazardous waste. More importantly, you recover valuable resources that put money back in your pocket.
But most recyclers are drowning in data chaos.

We recently worked with an e-waste recycler facing exactly this problem. They’d developed an innovative pyrolysis process that transforms e-waste into high-quality metal concentrates sold to smelters.
Sounds straightforward enough — but their operation was hamstrung by complexity at every turn.

Data Chaos in Three Dimensions

Purchasing: Flying Blind With Million-Dollar Decisions

The recycler purchases “lots” — shipping containers holding 10–20 tons of e-waste — from global suppliers. They don’t know what’s in these containers when they agree to buy them. Only after the shipment arrives can they inspect it to determine its actual content.

Meanwhile, pricing isn’t fixed — it’s based on complex contract bands tied to metal markets. For example, if copper content is below 10%, they pay nothing for it. Above 10%, they pay based on copper’s trading price as of the sampling date.

To make things trickier, these lots often get stored together, creating inventory management nightmares. Imagine buying mystery boxes worth hundreds of thousands, then piling them together before you’ve even checked what’s inside.

Production: Billions of Possible Combinations, One Excel Spreadsheet

The recycler can’t process individual lots efficiently. Instead, they need to mix multiple lots together before pyrolysis. But not just any mix will do. Each batch must:

• Meet minimum weight requirements (around 180 tons)
• Maintain plastic content within strict parameters (under 40%)
• Consider storage logistics (lots stored together need to be processed together)
• Maximize profitability based on complex downstream sales contracts
• How many combinations exist when mixing dozens of different lots with varying compositions? Billions. (That’s with a B, not an M.)

Their solution? Excel spreadsheets and “good enough” manual calculations. An operations manager would create mixes by hand, trying to find combinations that met requirements and seemed profitable. It worked — sort of.But “seems profitable” and “maximum possible profit” are worlds apart when you’re processing thousands of tons of material.

Sales: Downstream Complexity Mirroring Upstream Chaos

Sales contracts are just as messy as purchasing agreements. Multiple contract bands determine prices based on concentrate composition and market conditions. If a concentrate contains less than five grams of gold per ton, the customer might not pay for that gold at all. With such tight margins for error, every mixing decision impacts profitability. And the recycler was making these multi-million dollar decisions with the equivalent of a calculator and gut feeling.

Finding the Right Problem to Solve

For our proof of value project, we needed to identify the single highest-leverage problem we could solve quickly. After several weeks of engagement, we zeroed in on the mixing decision process.
The problem statement: Which lots should be mixed together for pyrolysis to maximize profit while meeting all operational constraints?

We picked this target for a reason: fix the mixing decisions, fix the profit margins. No need to overhaul their entire operation. Just get the mixes right, and the money follows.
From Static Guesswork to Dynamic Simulation Instead of relying on static spreadsheets and human calculation, we implemented a dynamic simulation engine built on Palantir Foundry.

The solution:

• Analyzes every lot in inventory or in transit
• Captures detailed composition data for each lot
• Runs hundreds of thousands of mixing simulations
• Calculates potential profit for each simulation scenario
• Filters results to find the highest-profit mix that meets all constraints

For example, the system simulates what happens when you mix Lot A with Lots B, C, and D, calculating the resulting plastic content, weight, and projected profit. Then it tries Lots A, C, D, and F. Then A, B, E, and G. And so on, through every viable combination.

A human with Excel can’t accomplish this. You might occasionally stumble on a good mix through manual calculation, but you’ll never consistently find the optimal one. The computational requirements are beyond human capacity. The system then presents the top mixing options that meet all constraints and yield the highest profit margin.

From Theory to Hard Numbers

To validate our solution, the recycler:

• Looked back at the previous month’s inventory
• Reviewed the mixes they had manually created and the resulting profit
• Loaded that same inventory data into our simulation tool
• Compared the tool’s recommended mixes against what they had actually done

The result? Our simulation identified mixing combinations that would have generated $300,000 more profit than their manual process — a 21.9% profit margin increase from a single month of operations.
And that’s not accounting for the hours of labor saved. Instead of painstakingly calculating possibilities in spreadsheets, operators now click a button and wait for results.

Beyond “Good Enough”

E-waste recycling flips traditional supply chains on their head. Suppliers call the shots, not customers. The companies collecting the junk have more leverage than processors, who have more power than smelters. It’s backwards from what you’d expect. This creates a situation where processors are pressured to accept material without complete information and make the best of what they receive. In this environment, “good enough” solutions like Excel spreadsheets become the default.

“Good enough” is only good enough for now. When you scale up, your problems don’t just grow — they multiply. The gap between “good enough” and “optimal” widens into a chasm that means the difference between profitability and loss. Every percentage point matters when you’re processing millions of dollars in materials monthly.

This recycler wasn’t profitable when we began working with them. The 21.9% margin improvement doesn’t just represent additional profit. It accelerates their path to break-even and sustainable operations.

Control Amid Chaos

We didn’t change this recycler’s business model. We just focused on one critical decision point: which lots to mix together. That single optimization led to a nearly 22% profit improvement. Now, this client has signed a 12-month contract with us to tackle additional challenges throughout their operation. The mixing optimization was just the beginning.

Do you run a recycling plant? A manufacturing line? Any operation with tons of variables and thin margins? Your data isn’t just sitting there — it’s hiding money. Real profits that you walk past every day. Excel gets you good enough. Data-driven simulation gets you optimal. The gap between those two points might just be the difference between thriving and merely surviving.

Turn your data chaos into business clarity. Better decisions. Bigger outcomes. That’s what we deliver at RANGR.