Magazine Store

50 Fastest Growing Companies of the Year 2023

We provide end-to-end solutions similar to other companies, but what sets us apart and contributes to our success is that we really focus on business value: Retief Gerber, CEO of Spatialedge


“We build only what is necessary, only when it is necessary, eliminating waste, and ultimately getting to market faster.”

Spatialedge is a big data and advanced analytics solutions provider, with a focus on enabling companies to rapidly build and operationalise robust analytical and machine learning models. It enables enterprise data science teams to deliver significantly more value to businesses faster and more reliably, by providing a toolkit consisting of engineers, data science specialists, open-source and bespoke software, and training programs.

Since its founding in 2015, Spatialedge has grown rapidly, capturing the South African market while the international expansion started in 2022.

The Silicon Review reached out to Retief Gerber, CEO of Spatialedge, and here’s what he had to say.

Interview Highlights

Q. What inspired the creation of Spatialedge?

Founded in 2015, the initial focus of our work was to determine what products and services we would build and offer within the data space as part of an aerospace incubator. Since the beginning, our company faced tight deadlines, difficult deliverables and needed to innovate within the broader data space. It was very important for us to figure out how to create value from data early on; otherwise, we would not have been able to survive.

In 2017, we refined our offering to focus on enterprise AI solutions using first-party data from large multinational companies. Our typical clients range from telecommunication companies, to retailers, financial institutions and mining companies. Our prior skills and knowledge gave us a competitive advantage when delivering business value to large companies by developing machine learning models and implementing them into robust production environments.

MLOps and machine learning engineering have been part of our practice even before they became widely used and accepted terms. This is because we have always been laser-focused on providing value to our clients and solving their problems.

Spatialedge has expanded from a team of two to a specialised group of close to 90 engineers, including software development specialists, data scientists, and engineers with expertise in machine learning, data engineering and Operations Research.

In essence, Spatialedge was created in response to the need for businesses to successfully solve challenging data and machine learning challenges, create successful products from their data, and run and manage these products in complex production environments.

Q. What methodology does Spatialedge implement to exceed client expectations?

We provide solutions for big data and artificial intelligence across cloud platforms and on-prem platforms. From the outside, Spatialedge appears to be a lot like other big data and artificial intelligence firms. We design and build MLOps pipelines, data pipelines, automated monitoring, metrics on our entire machine learning models, and continuous integration/continuous delivery pipelines. We develop data lakes, and data warehouses (or lake houses). Furthermore, we develop high-performance machine learning models and systems that enable the rapid roll-out of new competitor models. All depending on the current maturity of an organization and what will provide the most value to the organization,

In that sense, we provide end-to-end solutions just like other companies, but what sets us apart and contributes to our success is that we focus on business value. We build only what is necessary, only when it is necessary, eliminating waste, and ultimately getting to market faster. For example, we won’t do a “big build” such as building out a data lake for many years. Instead, we focus on the most valuable use case; develop the necessary models, datasets, and tooling to successfully deliver the value from that use case.

By relentlessly focusing on business value, having a high execution capability, and following lean and agile methodologies, we can solve difficult problems quickly. Solving real business problems directly improves our clients' bottom line.

Because we successfully solve these problems and increase the amount of money available to our clients, they use us to solve more and more problems. This helps them get ahead in their implementation and use of AI, giving them a massive competitive advantage.

Q. How does Spatialedge empower businesses and impact their decision-making process?

This depends on the use case, the business type, and their analytics maturity. To explain how we empower business through AI, it is useful to divide AI-driven decision-making into three categories: Planning, Operational, and Automated.

Planning: The first area affected by our AI decision-making tools is planning. Our tools empower planners with the necessary historic data, insights, forecasts, and scenario optimisation.

For example, retailers need to plan what products to buy, and how many of them need to be in which stores, by what date. In order for planners to create the best plans, they require the best available demand and sales forecasts. Retailers also need to plan and execute marketing campaigns and determine which product bundles to create. These tasks can be greatly improved with insights and forecasts derived from data.

In telecommunication companies, the tools help to plan changes and expansions to the network, such as where to build the next towers, and where to roll out additional infrastructure.

Operational: The next area is company operations. In retail, for instance, the products may not sell. To solve this problem, a retailer could discount these items, move them to stores that sell the same item better, or create promotions. The markdown optimiser identifies the optimal markdown strategy given a specific budget and the store transfer estimator automatically recommends what products to redistribute. This helps operational teams to make vastly improved day-to-day decisions.

Automated: Automated decisioning is for example when financial institutions need to make timely decisions on whether to provide credit to a new applicant or when a recommendation system is asked to provide personalized content for the company’s website.

Some systems are not fully automated and still require input from employees to develop strategies for handling edge cases, but with modern AI decision systems, a lot of the decisions are automated.

Q. In what industries are your clients? Can you provide us with one or two success stories describing the challenges your clients faced and how Spatialedge helped them overcome those challenges?

Retail: A retail engagement started with the problem of migrating a large data warehouse from an on-premise system to the cloud. From successfully managing that migration and adding multiple layers of improvement, such as data checks and monitoring to increase the robustness of the platform, we partnered up and got involved in building, productionising, delivering, monitoring and updating multiple use cases.

Noteworthy use cases include:

  • Developing a demand forecasting and planning system to better predict what customers want and deliver the products accordingly.
  • Developing a markdown optimisation tool that greatly improved the retailer’s sell-off rate from 75% to 93% while reducing the costs and duration of markdowns.
  • Successfully developing and deploying multiple decision applications, such as the basket analysis tool, retail experiments and a cannibalisation module being used by more than 150 employees in their decision-making.
  • Developing personalisation and recommendation systems.

As stated in 2022’s AI Africa Expo conference by the Head of Pepkor D&A, a total of R235 million in contributions towards profit has resulted in the investments that they have made.

Financial: We are engaging with multiple financial institutions, but our largest engagement started with a bank needing to get better visibility into their operations with less manual work. Since then, we have implemented multiple data pipelines, a massive data lake and continuously-updated web-based reporting used by management and the group CEO for his reports to the board.

Telecommunications: Our largest telecoms engagement started with the need for a team to develop hundreds of data pipelines, ingesting all their data (from many different sources and technologies) into multiple data lakes. The client also needed help in productionising and deploying their models on multiple, large, on-premise Hadoop systems.

Noteworthy projects include:

  • Building a fully hybrid (on-premise and in-cloud) data solution.
  • Building more than 700 data pipelines that are all monitored with anomaly detection and auto-healing.
  • Operationalising 100s of models, and reducing the time to roll out a model from prototype into production from +6 months to within a day.
  • Maturing their capabilities, with automated data and model testing and automated deployments.
  • Managing data ingestions in the excess of 178 billion records per day.

Q. What new endeavors is Spatialedge currently undertaking?

We are investing more than $4 million this year in expediting and expanding our Enterprise AI product offering.

Leadership | Spatialedge

Retief Gerber, CEO: Retief is a dedicated professional with extensive industry expertise. He has a ruthless approach to delivering business value to clients and partners. He is capable of identifying and implementing solutions to complex problems. His strategic vision and capacity to inspire teams have been important contributors to Spatialedge's success.

Dr. Frank Ortmann, COO: Frank has a Ph.D. in Operations Research and operates between the team management, business, and technical aspects of any data project. He is not scared to dive into areas which he is not familiar with and generally gets things done.


Dr. Jacques Du Toit, CTO: Jacques has incredible depth and insight into how machine learning, mathematics, and AI work. Even though he has his Ph.D. in machine learning, he is also highly skilled in operating many different computer systems and can dive deep into various issues.


Pierre le Roux, CMO: With experience building large-scale systems, combined with the love of solving customers’ problems, Pierre is an advisor to clients and helps them get their problems solved while driving the marketing and sales of Spatialedge.


Izaane Nortje, CFO: Izaane excels in an environment where ethical behavior and compliance are the cornerstones of business operations. She ensures transparent financial reporting and promotes financial sustainability that in essence optimizes growth.

“By relentlessly focusing on business value, having a high execution capability, and following lean and agile methodologies, we can solve difficult problems quickly. By solving real business problems we directly improve our clients' bottom line.”