Just about UX? Let’s not overlook the power of AI in the backend
- It’s long been a buzzword and, despite its challenges, the excitement has yet to dissipate
- But most applications of AI we hear about are centered on product and UX
- Here’s where AI is proving powerful backstage, according to Splunk
The buzz around artificial intelligence (AI) hasn’t gone anywhere in business. The tendency to attribute human-like characteristics to software is exciting, if only a reflection of our cognitive biases.
But beneath that initial fascination, AI has real substance in making use of the masses of data we are now accumulating. Simply, AI is machines executing tasks based on smart algorithms – it is computers learning from rich datasets and acting on them without being explicitly programmed to do so.
As consumers, or just people, we encounter this technology frequently in the form of predictive modelling, or machine learning, where models are built for making future decisions based on new data points. That takes form in the products and services we use daily, whether it’s Netflix recommending what you want to watch next, Google Maps knowing you’ll probably be going home at 6pm, or your Revolut account flagging an anomalous payment.
It’s the draw of smart products like these that have led to 93% of UK and US organizations considering AI to be a business priority according to a recent Vanson Bourne study commissioned by SnapLogic.
Too often, though, the power of AI and machine learning is enjoyed by the product user as UX benefits, and not necessarily by the product-makers, who may still be reliant on legacy solutions, among everyone else in the backend who have their own needs and requirements for the convenience that data-crunching algorithms can offer. Or for business-wide systems that could make enterprises more secure and resilient.
AI has serious applications in optimizing the backend workings of the business, and enabling organizations to predict and respond to trends, events and even threats, faster. Splunk’s AI and Machine Learning in Your Organization ebook sheds light on where AI is being used for full impact behind the scenes.
While an end-user product may be refined, the developers who made it may still be dealing with complex IT structures, thousands of alerts and an increasingly opaque environment.
Now, there are software systems that can autonomously improve and replace IT operations.
Artificial intelligence for IT operations, or AIOps, is the “marriage of big data and machine learning in IT”. Coined by Gartner in 2017, AIOps is now a growing trend in IT. It leverages historical data to boost productivity by allocating resources to low value, repetitive tasks, and enables the faster remediation of issues using a combination of predictive analytics and automated incident response.
As all businesses become tech-dependent and IT teams continue to grow, AIOps has become a burgeoning industry – with no shortage of vendors and specialists emerging – focused on performance monitoring, event correlation and analysis, IT service management and automation.
The end result? Time and money saved for the business and more productive (and happier) engineers. “Automation continues to be the most important end-goal for IT operations teams who are swimming in data and routine tasks,” OpsRamp SVP, Bhanu Singh, previously told TechHQ.
AI in cybersecurity
Gone are the days when businesses could ‘hide in the herd’; cybercriminal’s techniques are so far spread that just connecting to the internet opens the door to threats, including compromised websites, phishing emails, and distributed denial of service attacks.
Unfortunately, businesses are unprepared to fully prevent, detect, and respond to the growing number and sophistication of threats.
Consider that ransomware attacks occur every 14 seconds, according to a Cyber Security Ventures Official Annual Cybercrime Report. Given so many attacks, businesses are turning to AI and machine learning capabilities to help shore up a scarcity of cybersecurity experts.
In cybersecurity, machine learning has applications in advanced threat detection and stopping insider threats, which require a more nuanced approach to monitoring and response. Sophisticated attacks that move laterally within a network, or breaches caused by unwitting access to sensitive information can be tackled by automated and intelligent anomaly detection.
AI and machine learning can enable analysts and security teams to paw through masses of log and event data from applications, endpoints and network devices to conduct rapid investigations and uncover patterns to determine the root cause of incidents.
As the threat landscape evolves, and the cost of a cybersecurity breach becomes increasingly catastrophic for small and large businesses alike, AI and machine learning is handing organizations improvements in detection speed, impact analysis and response.
Data, and lots of it, is core to the success of any AI or machine learning initiative. To leverage the benefits of these intelligent systems within the organization, businesses must be prepared to conduct the manual effort and resource required to refine large volumes of data that give AI models the fuel they need to learn and burn.
At IBM — a company with a better view than most of the emerging technologies market — data-related struggles are a top reason the company’s clients have ceased or cancelled AI projects, according to the firm’s SVP of Cloud and Cognitive Software, Arvind Krishna.
Speaking at Wall Street Journal’s Future of Everything Festival last year, Krishna said that companies are finding themselves underprepared for the work and cost of acquiring and preparing that data— work comprising about 80% of an AI project.
“[…] you run out of patience along the way, because you spend your first year just collecting and cleansing the data,” said Krishna. Companies can become impatient and disillusioned with the work, he explained, and “kind of bail on it.”
However, the more data you have, the better, and once the grunt work and heavy-lifting is out the way, effective AI and machine learning means organizations are no longer bogged down by data; they are elevated by it. The challenge is getting there, but the benefits speak (and work) for themselves.