WeBuyCars: How an R25 Billion South African Business Uses AI to Buy Cars Autonomously

// Chapter 01 — The company hiding in plain sight

Most South Africans think of WeBuyCars as a place to sell your car quickly. You drive in, they inspect it, they offer you a number, you take it or leave it. Simple. Physical. Human.

That description is accurate for the customer experience. It completely misses what is actually happening underneath.

WeBuyCars is an R25–R30 billion revenue business. Last year it bought 180,576 vehicles and sold 179,006. That is not a car dealership with a good location. That is a high-throughput inventory-turn machine operating at a scale that requires technology to function at all. And increasingly, that technology is AI.

According to chief digital officer Wynand Beukes, WeBuyCars runs two major AI systems internally — Blue and Orange. Orange is the customer-facing large language model on the website. Blue is a collection of machine learning models that hold pricing and vehicle information, based on historical data, purchase and sale history, and market trends.

Here is the number that should stop you: Blue has bought just over 2,800 cars autonomously, without any human pricing involved. An AI model is making purchase decisions on used vehicles. At scale. Without a human in the loop.

That is what a tech company looks like when it is wearing a car dealership's clothes.

"Do not build technology for its own sake. Focus on technology that improves unit economics." — Dr Wynand Beukes, Deputy CEO, WeBuyCars, NADA Connect 2026

The WeBuyCars technology stack — AI, computer vision, and Blue AI explained

// Chapter 02 — What they actually built

WeBuyCars has accelerated the rollout of Inspectify — its in-house vehicle inspection platform — using richer vehicle data and advanced pricing models to sharpen valuations. Inspectify is not a third-party tool they bought. They built it themselves. It employs trained technicians, strict quality control procedures, and independently audited processes to standardise every inspection across every branch.

Computer vision models evaluate images of cars, detect damage, and analyse undercarriage photos made from thousands of stitched images. These models feed directly back into pricing accuracy. The person doing the inspection is still human. The intelligence interpreting what they see is not.

Generative AI supports operations ranging from customer conversations to automated social media responses. On the website, an AI assistant named Orange can explain vehicle features, compare models, and even calculate the cost of driving between cities.

This is a company that started as two brothers personally inspecting cars and parking stock on borrowed lots. Today it has 129 branches and buying pods across all nine provinces, with plans to reach 200. The technology is what makes that scale possible without chaos.

// WeBuyCars by the numbers · 2026

R25B+

Annual revenue

180K

Vehicles bought per year

2,800

Cars bought by AI autonomously

129

Branches & buying pods

What WeBuyCars' AI strategy means for your South African business

// Chapter 03 — The translation

You are not running a used-car empire. You do not have a data science team or a chief digital officer. The WeBuyCars story can feel like it belongs to a different category of business entirely.

But the principle underneath it is not complicated, and it applies to every business that buys and sells anything — cars, property, second-hand goods, services, inventory of any kind.

The principle is this: the business that knows its data best wins the pricing game. WeBuyCars built Blue because pricing a used car accurately, thousands of times per week, is a data problem. The human who has been buying cars for twenty years has intuition. The machine that has processed 180,000 transactions has data. At scale, data beats intuition.

You do not need Blue. But you do need to ask: where in my business is pricing being done by gut feel that could be informed by data? Where is the customer experience being handled by a human for a task that a well-configured AI could handle at 2am on a Sunday? Where is my inspection, assessment, or valuation process inconsistent because it depends on who is doing it that day?

Those are the WeBuyCars questions. And they apply whether you are running a panel shop, a guesthouse, a cleaning company, or a consultancy.

Building something similar? Start with the right infrastructure.

If you are thinking about adding data-driven pricing, AI-assisted customer intake, or automated valuation to your business, the architecture underneath it matters more than the AI model you choose. Forge Vertical builds the infrastructure first — the database design, the API architecture, the security layer — and then wires in the AI. That is the order that works.

Tell us what you're building →

The lessons — applied

// Chapter 04 — Six things WeBuyCars got right that most businesses ignore

// Lesson 01

Technology is not a department. It is the business.

WeBuyCars did not build a tech team to support the car business. The car business IS the tech business. Wynand Beukes — the CDO — is also the Deputy CEO. That is not an accident. The technology leader is the business leader. Most businesses have IT as a cost centre. The ones that win have it as a value centre.

// Lesson 02

Build your own tools when the problem is core to your business.

Inspectify exists because vehicle inspection is core to WeBuyCars' unit economics. They could have bought a third-party inspection tool. Instead they built one that feeds directly into their pricing models. If a process is core to how you make money, building your own tool gives you a data advantage that buying never will.

// Lesson 03

AI on pricing is the highest-leverage place to start.

Blue does not replace the car buyer. It gives the car buyer — and increasingly, replaces the car buyer — a pricing recommendation grounded in 180,000 transactions of historical data. Every business that buys or sells goods has a pricing problem. Start there. The ROI is immediate and measurable.

// Lesson 04

Customer-facing AI buys you time at scale.

Orange handles questions about vehicle features, model comparisons, and even driving costs between cities. That is a customer service operation that would require staff at every hour of the day. Instead it is one AI model, available at 2am, consistent across every interaction. The human staff handle what requires human judgment. Orange handles the rest.

// Lesson 05

Standardisation before automation.

Inspectify standardised the inspection process first — trained technicians, audited procedures — before the AI models interpreted the results. You cannot automate a chaotic process. You get automated chaos. The businesses that fail with AI almost always skip this step.

// Lesson 06

Technology is how you expand the market, not just compete in it.

WeBuyCars plans to expand from 129 to 200 branches and buying pods. That expansion is only possible because the technology makes every new location plug-and-play. The pricing model works in Polokwane the same way it works in Cape Town. The inspection process is the same. The AI does not get tired or inconsistent. Technology is not just efficiency — it is the mechanism of geographic and market expansion.

The honest question for every business owner

// Chapter 05 — What are you actually building?

WeBuyCars started as two brothers personally inspecting cars. The idea was strikingly simple at the time: create a fast, hassle-free way for people to sell their cars for cash, avoiding classifieds, private buyers and dealership trade-ins. The technology came later, built deliberately to solve real operational problems as the business scaled.

That sequence matters. They did not start with AI. They started with a problem worth solving and customers willing to pay to have it solved. The technology was the answer to the question: how do we do this at scale without losing quality or margin?

The question for your business is not "should I use AI?" The question is: what problem in my business is worth solving at scale? What process, if it were faster, more consistent, or more data-driven, would directly improve how much money the business makes or how much time it saves?

That is where the technology conversation starts. And that is the conversation Forge Vertical builds from.

Aurora Repair — the same principle, applied to panel beating: Aurora Repair (aurora-repair.com) takes photos of vehicle damage and returns an itemised repair assessment — the same computer vision approach WeBuyCars uses for valuation, applied to the repair side of the automotive market. Built in partnership with International Panel Shop, Cape Town. The same principle. A different application. A different market.

// Want to add a technology layer to your business?

Start with a brief. Get a scope and price in 24 hours.

Whether you need a data-driven pricing tool, an AI customer intake system, or a complete platform rebuild — Forge Vertical builds the infrastructure first and the AI on top of it.

Start a project → WhatsApp Jarrit →

Written by

Jarrit Hosking

Forge Vertical · Cape Town

← All articles Aurora Repair →