Subscribe
About

How to build a modern, unified data and analytics platform

By Kirsten Doyle

Data is skyrocketing and companies the world over are rushing to put themselves in a position to take full advantage of the insights and benefits that a modern toolset with AI and ML can add to their organisation.

However, there are several challenges associated with running a traditional enterprise data warehouse, which is why moving it to a modern cloud environment is key.

With this in mind, Thomas Fowler, CTO at CloudSmiths, and Louis van Schalkwyk, technical operations, at Digicloud Africa, will be presenting on “The benefits of building a modern, unified data and analytics platform”, at The ITWeb Cloud & Data Centre Summit 2022 to be held at The Capital on the Park, on 1 November.

ITWeb Cloud & Data Centre Summit 2022

This year’s event will examine the relationship between data centres, cloud and DevOps in accelerating change, with several international and local experts presenting on a wide range of enterprise-critical topics, such as the state of data centres in Africa, simplifying cloud adoption journeys, multi-cloud, enterprise agility, DevSecOps, and many more. For more information and to register, click here.

According to Van Schalkwyk, getting value from data takes a multifaceted approach and one that requires a multitude of tools, services, and skills to get right. “Data is all over the place, often siloed, incomplete, and out of date. Not all data is accessible from a single place. All these factors make it difficult to get accurate insights from your data.”

When it comes to building a modern, unified platform, he says customers need to ensure they have robust data pipelines in place that can handle disruptions, scale automatically and take full advantage of managed services offered by cloud providers.

Then, says Van Schalkwyk, they need to ensure they choose the right data warehouse solution that will meet their future aspirations, like machine learning. “Knowing that the enterprise data warehouse they choose can scale with them, and will provide the tools needed over the next five to 10 years is crucial.”

There are also several pitfalls to avoid, he explains. “Avoid getting locked into systems with complex and proprietary licensing models. A lot of services out there are easy to get started with but become very expensive when you need to scale out. Also, use systems that can talk to and be integrated with other systems, and avoid services that require manual management of clusters and servers as far as possible.”

Delegates attending their talk will hear how other SA companies have moved analytic workloads to Google Cloud, leveraging Vertex AI as a single platform to manage data workloads, from loading, and transforming, all the way to serving ML models hosted on Google Cloud.

Share