Mavenir Collaborates with Red Hat to Deliver Carrier-Grade On-Premise Conversational AI and Agentic AI Service Assurance for Telcos

February 18, 2026,

Mavenir announced a collaboration with Red Hat to deliver Conversational AI and Agentic AI service assurance solutions using Red Hat AI. These carrier-grade generative AI (gen AI) offerings are powered by Red Hat OpenShift AI running on-premise, empowering telecommunications service providers with a highly scalable, security-focused, resilient and cost-optimized foundation for network operations.

This collaboration is focused on transforming the telecommunications industry by providing service providers with AI solutions that leverage their existing platform investments while delivering additional security, resiliency, with low latency that is critical for real-time network operations.  

With this effort, Mavenir provides AI solutions tailored to the telecommunications industry that are purpose built to be more cost-effective and more secure than hosted as-a-service AI solutions.

Sachin Karkala, EVP & General Manager, IMS & RAN of Mavenir, stated: "In the race to monetize AI and evolve to autonomous networks, service providers need control of the user experience, cost and data privacy. Our collaboration with Red Hat enables us to develop on-premise AI solutions that do just that. Generative AI models can be fine-tuned using service provider data to build specialized, on-premise agents helping deliver optimal performance.”

With conversational AI, the industry is moving up the value chain from merely providing connectivity between entities, to AI agents taking over critical control of acting on the content and intent of the communication.  

Agentic service assurance is a significant evolution of today’s service assurance solutions relying on machine learning models alone. Agentic systems have a broader perspective, being able to create workflows and drive decision making between different domains in the network and disparate data sources. They adapt and self-learn, ultimately being able to detect and correct issues before they impact the customer.

As a step toward deployment readiness, Mavenir and Red Hat engaged with the Dell Technologies Open Telecom Ecosystem Lab (OTEL) which hosts an advanced AI infrastructure based on the Dell PowerEdge R760xa – a high-performance air-cooled server engineered for a wide range of applications, including AI, machine learning and deep learning-training and inferencing. Dell’s OTEL AI experts collaborated with the Mavenir and Red Hat engineering teams to fine tune the large language model and maximize the solution’s efficient use of the underlying infrastructure.

By deploying these solutions on-premise with Red Hat OpenShift AI, service providers can optimize the efficiency of their GPU resources, maintain full control of security and governance requirements and more easily manage the complex AI lifecycle.

Honoré LaBourdette, Vice President of the Global Telco Partner Ecosystem at Red Hat, commented: "For enterprise AI to move from proof-of-concept to production, it requires a validated foundation with mandatory components like guardrails, observability and a robust lifecycle. Red Hat OpenShift AI provides that necessary foundation. Together with  Mavenir, we are demonstrating cost-effective GPU efficiency through tools like Kueue and Dynamic Accelerator Slicer (DAS) for multi-tenant resource sharing, which helps optimize utilization of high-dollar hardware investments. This, paired with low-latency model serving using vLLM, helps ensure that real-time applications like AI voice translation, achieve the deterministic performance the market demands, proving the tangible value of the on-premise model. This is our first announcement in a series of joint initiatives with Mavenir to leverage Red Hat AI technology."