RADCOM (NASDAQ: RDCM) says it is continuing investment in its ACE platform and agentic AI roadmap while adding VNPT and CETIN Networks as customers, linking its network-assurance strategy to cloud-native deployments that are intended to spot service issues before they reach subscribers. The update matters to operations teams because it puts anomaly detection, customer-impact context and multi-cloud deployment at the center of an assurance platform rather than treating monitoring as a separate reporting layer.
The company disclosed the plans with third-quarter results released November 11. RADCOM reported revenue of $14.0 million, up 19.4% from the preceding quarter but down 23.6% from a year earlier. It returned to non-GAAP operating profitability, reporting $0.7 million in non-GAAP operating income, compared with $3.8 million in the 2025 quarter. The financial result is not a product launch, but the customer and investment details offer a current view of where RADCOM is placing its AIOps development effort.
VNPT, a Vietnamese telecommunications operator with more than 30 million subscribers, selected RADCOM to provide AI-native, cloud-native service assurance for its roaming network, according to the company. RADCOM said the work will replace an incumbent provider after a competitive process and deploy the ACE platform across multiple cloud environments. The company has not provided a production timetable, commercial value, rollout scope or performance targets for the deployment.
RADCOM ACE is designed to provide real-time, subscriber-aware visibility across network services. The company said the platform applies artificial intelligence and machine learning-based anomaly detection to surface issues before they affect customer experience. In practical terms, that approach aims to join network signals with an assessment of the users or services exposed to a fault, a task that becomes more difficult when traffic, infrastructure and operations systems span several clouds.
For network operations teams, the key distinction is between collecting alarms and deciding which condition should receive attention first. A roaming issue can involve radio, packet-core, policy, partner-network and service components. A cloud-native assurance platform may consolidate some of that telemetry and context, but it does not eliminate the need for teams to validate the suspected cause, coordinate changes with partner networks and confirm that a remediation worked.
RADCOM also said it is investing in an agentic AI roadmap. The company did not describe the future capabilities, delivery date or decision authority that its agents will have. That omission is material. An agent used to summarize an incident or recommend a correlation rule presents different operational and governance questions from one that can change policy, reroute traffic or open a service ticket without approval. Enterprises evaluating the roadmap will need to distinguish present ACE functionality from forward-looking AI plans.
The company reaffirmed its revised full-year 2026 revenue outlook of $57 million to $63 million and reported $111.4 million in cash with no debt. Those figures explain its stated ability to fund product work, but they do not establish that planned AI functions will improve network availability or reduce operating costs. Those outcomes will depend on deployment quality, telemetry coverage and the processes operators use to act on platform findings.
The broader operational signal is that service assurance vendors are moving toward platforms that combine multi-cloud visibility with AI-assisted analysis. The older approach of operating isolated probes and reviewing large queues of alerts can leave teams to reconstruct a customer-impact problem after it has spread. RADCOM’s VNPT deployment is a concrete example of the newer model, though the practical value will be determined by the depth of its integration, the accuracy of its detections and the guardrails around any future agent-driven action.