Google Cloud today announced the public launch of its Network Intelligence Center, a platform that uses AI to help manage the sprawling networks companies must wrangle as they move more of their critical operations online.

As cloud computing evolves, companies increasingly have several cloud services, and in some cases hybrid ones that blend premise-based services with virtual ones. While promising efficiencies and enabling new services for customers, this tangle also creates massive headaches for IT managers.

According to Shailesh Shukla, vice president of product management for Google Cloud Platform Networking, the tools to oversee this patchwork have lagged behind, leaving the networks vulnerable to simplistic errors such as misconfigurations that can take the whole system down.

“Adoption of hybrid and multi-cloud is absolutely critical for organizations to remain agile,” Shukla wrote in a blog post. “However, this underscores the need for intelligent and continuous network operations — the promise that the network is doing what it needs to do, in line with business intent.”

Google’s Network Intelligence Center is designed with two broad goals: Bring greater automation through artificial intelligence and machine learning; and be proactive about finding and fixing issues.

While the center has been in beta, it is now publicly available with four modules that are in various stages of development:

  • Connectivity Tests: Currently in beta, this module diagnoses issues related to connectivity while also testing any changes in advance.
  • Network Topology: Currently in beta, this module visualizes all networks and how they interact to make it easy to understand and monitor the global infrastructure.
  • Performance Dashboard: Currently in alpha, it offers real-time network performance monitoring to keep tabs on the overall health and quickly spot any application performance issues.
  • Firewall Metrics & Insights: Currently in alpha, this module gives greater visibility into security metrics of firewalls, and allows faster responses to adapt rules when necessary.

The company says it’s also developing several other modules that will be available for testing in the coming months.