Don’t forget the end user in your AI cloud strategy
When you’re setting up your business to make the most of artificial intelligence (AI), you’ll need to consider connectivity to the cloud, simply because that’s where the majority of AI apps are hosted. You might begin with connectivity from your on-premise infrastructure to the cloud. You might also have considered a cloud routing service so AI data flows seamlessly between cloud environments. But what about the connection from the cloud to the end user?
The recipients of outputs from AI applications aren’t necessarily in the cloud. To reach them, businesses need flexible, robust cloud connectivity across all sites and remote-worker locations, as well as customer-facing apps and other end-user digital products.
Here are some different types of end users and how to factor them into your business’s AI strategy.
Who is the end user of AI?
The most obvious end users to consider in your AI strategy are humans. These include the business’s internal teams and decision-makers who are increasingly likely to be working remotely, as well as field workers such as engineers. Then there are your customers who might be accessing AI-based insights or recommendations.
On the non-human side sit the apps and AI-powered machines or robots doing everything from handling customer queries via chatbots, through to roles in production, manufacturing and construction. They may even be involved in higher-stakes functions such as surgery or autonomous vehicles.
End-user interactions with AI data and applications vary by industry. For example:
- Automotive – drivers use AI-informed sensors to take real-time mitigating action that avoids hazards or conserves fuel and battery life. They can also use AI-powered route recommendations based on real-time traffic data.
- Ecommerce – apps and websites draw on AI stock insights to provide live delivery times. They also provide AI-driven product and sizing recommendations based on user data that is cross-referenced with trends from reviews and returns.
- Financial services – consultants and risk analysts use performance data and metrics across the business to give advice based on AI algorithms and trend analytics. And banking customers receive AI-powered financial product recommendations and budgeting suggestions from their data.
- Recruitment – portals search opportunities and automatically match candidate profiles, or recommend job adverts to applicants in real time.
Getting end-user connectivity right for AI
For this busy ecosystem of cloud-based AI applications and data to run efficiently for all end users, the connections need to be fast, secure, and reliable. High latency in cloud connectivity can slow operations, cause intermittent service or lead to errors in AI performance. All this means the business’s AI investment goes to waste or returns limited value.
Achieving minimized latency and maximized performance with AI applications requires short, direct pathways from the cloud to the end user. To put that into perspective, applications for self-driving cars need to be less than 80km from the user, equating to 1 millisecond of lag.
There’s also security to think about. Cloud connectivity needs to be robust to protect data as it travels between in-cloud AI applications and the end user to minimize the risk from cyber attacks. Resilience is also vital to maintain access to AI applications and reduce downtime.
End-user connectivity will become even more important with the emergence of immersive tech and virtual reality devices that rely on a finely calibrated interplay of synchronized video, audio, and sensory data. At the beginning of the year, for example, Apple released its VR device which uses AI and machine learning to relay users’ applications and data from the cloud into virtual spaces. These applications can fall down in as little as half a millisecond of lag, since that’s how long it takes the human mind to perceive delays in sensory input. So AI has almost literally no time for weak connectivity.
What else shouldn’t you forget when it comes to AI in the cloud?
From avoiding the risks of cloud concentration to finding data centers with the power density for AI servers, there are several key considerations for connectivity that makes AI work effectively in the cloud. Find out what they are and why they’re so important in our guide, The secret to AI success.