What challenges did Ericsson face before AI implementation?

Communications service providers face rising complexity and operational cost managing multi-technology 5G/6G environments. Manual interventions make it difficult to scale network operations and slow the move toward autonomous networks. Legacy systems hinder adoption of cloud-native AI automation.

What was the impact of the AI solution?

The article cites 98% field-validated accuracy. It reports 54% faster cell issue resolution. It reports 75% reduction in network optimization time and effort. It reports 43% improved downlink throughput in problematic cells and over 4% spectral efficiency gains. It says the solution is deployed across more than 60 CSPs, managing 13+ million sites and serving 2+ billion subscribers.

Rank#35of 3,117

Source

Ericsson agentic rApp as a Service on AWS for autonomous RAN optimization

Use case typeAutomotive operations automationUpdated Jun 13, 2026

Ericsson describes an AWS-hosted rApp as a Service solution that uses agentic AI to automate RAN operations and optimization for communications service providers. The platform integrates with Ericsson Intelligent Automation Platform and O-RAN interfaces to coordinate specialized agents for anomaly detection, root-cause explainability, cell shaping, and uplink performance optimization. It is delivered as a multi-tenant SaaS offering through AWS Marketplace.

Organization: Ericsson
Industry: Tech & Comms
Location: Sweden
Published: February 2026

Reported outcomes

+98%

field-validated accuracyQuality & accuracy

+54%cell issue resolution speed−75%network optimization time and effort+43%downlink throughput in problematic cells+4%spectral efficiency gains

Strategic outcomes

New business modelDelivered as multi-tenant SaaS offeringNew product / capabilityAutomated RAN operations and optimizationScale & capacityDeployed across major CSP network scale

Catalog median for quality & accuracy deployments: +90% across 287 reported metrics. Compare benchmarks →

Primary read

Use case focus

Showing 3 of 4

1Agentic AI
2AIOps
3Network Optimization

Communications service providers face rising complexity and operational cost managing multi-technology 5G/6G environments.
Manual interventions make it difficult to scale network operations and slow the move toward autonomous networks.
Legacy systems hinder adoption of cloud-native AI automation.

Ericsson rApp as a Service is hosted on AWS and uses Amazon Bedrock agent capabilities for the agentic AI layer.
The solution uses Amazon SageMaker AI, Amazon ECS on AWS Fargate, AWS Glue, Amazon Athena, and AWS Lambda for ML, orchestration, analytics, ETL, and automation.
The platform coordinates multiple specialized AI agents through a supervisor-agent pattern and integrates with Ericsson Intelligent Automation Platform via O1/R1 interfaces.

Technologies

Amazon Bedrock Amazon SageMaker AI Amazon ECS on AWS Fargate AWS Glue Amazon Athena AWS Lambda AWS Marketplace Ericsson Intelligent Automation Platform O-RAN R1 O-RAN O1

The article cites 98% field-validated accuracy.
It reports 54% faster cell issue resolution.
It reports 75% reduction in network optimization time and effort.
It reports 43% improved downlink throughput in problematic cells and over 4% spectral efficiency gains.
It says the solution is deployed across more than 60 CSPs, managing 13+ million sites and serving 2+ billion subscribers.

Architecture

Ericsson rApp as a Service is a SaaS solution hosted on AWS and available via AWS Marketplace. It integrates with Ericsson Intelligent Automation Platform (EIAP) and O-RAN O1/R1 interfaces. The architecture uses a supervisor agent coordinating specialized rApp agents for telemetry-driven network optimization. AWS services named in the article include Amazon Bedrock agent capabilities, Amazon SageMaker AI, Amazon ECS on AWS Fargate, AWS Glue, Amazon Athena, and AWS Lambda. The platform is multi-tenant and uses standardized interfaces and APIs/MCP/A2A for integration.

Sources & evidence1

Groundedness: 5/5Type: Blog PostPublished: Feb 17, 2026Publisher: EricssonEvidence: VendorConfidence: Medium

Primary source

AI-generated summary. Verify important details with the linked sources before relying on this case.

Explore related AI use cases

Customer evidence