The IT press is awash with technology terms that people use without really understanding. Vendors often put their own spin on these terms to highlight their own products' capabilities. Here are simple explanations for five technology phrases that you'll run across at every technology conference.
Even washing machines have artificial intelligence these days, but what it is it and why is it important? Traditionally, computers could only follow procedural steps with clearly defined outcomes. They were not good at tasks that required judgement. Only a human could decide whether a car door had sustained damage by examining a picture, or spot unusual patterns in financial transactions.
AI enables computers to handle those judgement-based tasks. It usually involves training the computer using many examples of what it is trying to find, (for example, thousands of pictures of damaged car doors), in a process called machine learning. More recently, companies have ventured into unsupervised learning, where the computer looks at lots of data and spots patterns on its own.
Big data is information that is so large that it is difficult to process using normal tools like relational databases. It also grows quickly over time. Dealing with big data is like drinking from an information firehouse - the stuff just keeps coming, and you have to find a way to store and use it effectively.
Big data often includes both structured data, like customer transactions, and unstructured data, like social media posts. It's a great resource for finding trends in your business using analytics. It's also a useful source of AI training data.
One way to structure big data is in a data lake, which is a large data store that keeps all of the information as-is and allows you to process it for specific tasks later. You'll typically offload big data processing to a technology like Hadoop, which cuts it into smaller chunks and processes them concurrently to get faster results.
If you create your own software, then you'll eventually run into DevOps. It's a discipline for streamlining software development and deployment by joining the two together more effectively. It shortens the time between coding a piece of software and deploying it in the field, and it enables developers to act more quickly on user feedback to create more frequent software updates.
DevOps achieves this by encouraging developers to share ownership of the IT resources that their software uses. They can often provision these resources themselves in code, conjuring up a virtual machine and virtual storage to support their applications.
DevOps uses underlying technologies to streamline processes. This automation software is often in the cloud, and creates pipelines that funnel code from creation to deployment while imposing strict quality controls. This is known as continuous integration and continuous deployment (CI/CD).
Digital transformation is the adoption of digital technology to fundamentally change the way a company does business from the inside out. Not only does it digitise manual processes inside the company, but it introduces new technologies that enable the company to deliver employee and customer experiences that weren't possible before.
Digital transformation could enable an insurance company to move from claims processing based on phone calls and site visits to submission of images via a mobile phone-based system. A used car dealership might expand its reach using an online marketplace that allows people to submit vehicles for sale, matching them with potential buyers.
Digital transformation projects involve more than just technology. Companies must set clear goals and often alter their organisational structures to support them, creating new responsibilities. They must often also set new metrics to measure project successes and tie them to clearly defined goals.
Over the last few years, IT operations have virtualised many resources including servers, desktops, and storage. They've used software to share hardware resources, so a single physical server might look like ten servers to users. They have also redefined networks in software, making them easier to configure.
A software-defined wide area network (SD-WAN) applies that virtualisation to wide-area networks spanning whole countries and beyond. It decouples the underlying network routing hardware from the control layer, which configures functions like packet management and routing. The result is a network that you can quickly configure from a central location to evolve with your business.
Traditionally, corporate WANs use dedicated carrier services like multi-protocol label switching (MPLS). This kind of network connection offers guaranteed performance and security direct to a location but can be expensive and time-consuming to set up.
SD-WAN services replace the expensive MPLS routing equipment at remote locations with simple virtualised solutions. These can communicate over conventional broadband, but they create a connection that is better than a standard consumer broadband service. That's because the central software-based control allows admins to configure properties like encryption, quality of service, and network routing on a per-user and per-application basis.
A retailer might use SD-WAN capabilities to quickly bring new locations online. A company with remote workers could set up network support for home offices. Software configuration allows this to happen far more quickly than it would if working through a traditional carrier offering dedicated WAN services like MPLS, making remote locations far more agile. Most importantly, it lets people connect remote locations far less expensively than they could using MPLS.
Don't be bamboozled by vendor rhetoric. The next time someone brings up one of these terms, you'll be in a better position to understand their speil - and decide whether they know what they're talking about.