We’ve written before about how AI and automation can combine to drive greater efficiencies. In a recent Datamation article, Drew Robb explores a similar theme, outlining new AI automation trends emerging in the second half of the year, among them:

1. AI Training and Quality 

As those of us in the industry know, real-world data is rarely neat and predictable. Quality issues such as anomalies, unexpected duplications, missing, truncated, or invalid data can all lead to poor decision-making with the potential to negatively impact the bottom line. ML and AI algorithms that rely on making inferences and adjustments via incoming data are particularly vulnerable to bad data conditions. Automation can address this, enabling engineers to automate discovery by running pipeline tests on incoming data and spotting outliers before they make it to the AI training process.

2. AI Automation and Networking 

Network operations have traditionally involved a significant amount of human oversight and effort. Most of this effort is manual and exploratory, with engineers attempting to determine the source of problems without much insight or data ahead of time. Today, there is an opportunity to leverage AI automation for analytics and data, and automate much of the network operations effort. For example, AI automation can perform root cause analysis, spot problematics sites, devices, and protocols, identify supporting evidence, and provide remediation actions to network engineers.

3. AI in Network Security

AI’s ability to spot hidden details can be instrumental in security. For example, AI is being used to automate baseline profiling and anomaly detection so companies can detect zero-day attacks. In addition, the technology can rank security alerts and suggest corrective actions so that response teams can be nimbler and more efficient.

4. Process Automation 

AI-driven automation can perform numerous logistical or remediation tasks without requiring human intervention. Because these systems continuously and autonomously learn, they are continually improving, and companies can also customize them to address specific needs or requirements.

5. AI as a Service 

As the technology matures, AI as a Service is emerging as an attractive option for companies that either don’t want or can’t afford to handle the complexities of AI engines and platforms.

Robb has more on this and the aforementioned trends via Datamation.