About OAI Workshop
OpenAirInterface summer workshop was held on May 21-22 and the event was hosted by Ericsson at its prestigious headquarters in Kista, Sweden. This event is the yearly gathering of OAI experts, developers, researchers, and the broader OAI community
Defining the technical capabilities required for an AI-native 6G reference, implementation is critical to shaping OAI’s roadmap and reinforcing its role as a key enabler for future wireless networks. This workshop provided, once again, a platform for experts from both academia and industry to exchange insights, share technical advancements, and help steer OAI’s development in response to these evolving needs.
During the two workshop days, we explored the needs we expect OAI to fulfill for 6G and AI-driven networks and which BubbleRAN is excited to be a part of. On the day 2, BubbleRAN participated to a panel session “How are Businesses Leveraging OAI and other Open Source: Markets and Opportunities”. Discussion included Navid Nikaien, BubbleRAN CEO, together with Paulo Marques Allbesmart , Arpad Iozsa Lasting Software, Tommaso Melodia zTouch and Laurent Thomas Open Cells Project.
We truly believe that pioneering research and open-source code contributions are keys to build an open 5G+ ecosystem enabling the freedom and diversity of choice. Engaging discussions, we warmly thank all participants.
SUMMARY VIDEO
Our Innovative Demo
This year we showcased MX-AI: Intent-Driven RAN Automation Powered by LLM
Our live demonstration showcased MX-AI, BubbleRAN’s Open and modular ecosystem and development kits designed to accelerate the Development, Deployment, Orchestration of intelligent RAN automation across multi-vendor, multi-domain, and multi-environment (multi-x) networks.
MX-AI brings together users, developers, and AI agents in a unified and cloud-native platform that enables seamless testing, integration, and deployment of xApps, rApps, AI agents, and model into production network, all aligned with 3GPP and O-RAN standards.
In this demo, we highlight intent-based RAN control using a Large Language Model (LLM) agent built with the ADK and rApp SDK. The LLM agent interprets high-level intents and translates them into network policies. These policies are then dispatched to appropriate xApps which dynamically enforce them within the RAN.
We demonstrate this workflow using AutoRAN, a pair of intelligent rApp and xApp that adaptively respond to network conditions in real time based on dynamically generated policies. A continuous observability loop provides full performance visibility and feedback, allowing the system to monitor:
- Outcomes
- Refine decisions
- Improve future policy responses over time
The entire process, from expressing an intent to executing real-time actions, is fully automated with minimal manual intervention. Watch the video to see the demo.
BubbleRAN Ecosystem for Multi-X Automation & Intelligence
Are you interested to know more about our MX-AI?
- Yes > contact us using the quick contact form
- Yes > You can also Book a meeting with us
- Yes > visit product page