Meta XR over Verizon MEC infrastructure

This initiative aimed to assess network performance for Meta's AR/VR applications over Verizon's 5G network. The evaluation focused on three transport technologies: Direct Connect, Internet-based communication, and BGP AS peering. Key performance indicators (KPIs) analyzed included jitter, latency, throughput, packet loss, and retransmissions.

Achievements:

• Evaluation of QCI133 for real-time, low-latency applications.
• Jitter and Latency improvement by 10% over VzW Wavelength.
• Test ecosystem development over containers (Docker) at scale.

L4S Evaluation Meta, Apple, XBox

Building on the evolution of Meta's XR initiative, this project aimed to assess the use and functionality of L4S over Verizon’s 5G network. The goal was to detect traffic congestion before it impacted user experience and leverage AI/ML frameworks to predict and optimize packet flow by adjusting network queuing and data compression rates.

Achievements:

• Evaluation of QCI133 API call on demand.
• Full-Stack ecosystem deployment at scale, productivity improvement.
• Developed AI/ML predictive models for network automation improvement.
• Resulted in a patent submission for network performance improvement by using AI/ML prediction models.

OpenSesame HiFi Audio on realtime

OpenSesame is a high-fidelity real-time audio collaboration platform that enables musicians to collaborate remotely. This initiative focused on evaluating and integrating the low-throughput/low-latency API as an automated solution in two key scenarios: as an on-demand trigger during network congestion or as a default end-to-end configuration applied at the start of the traffic flow.

> OpenSesame
> Verizon, OpenSesame, and AWS Collaboration
> Architecting multi-carrier interoperability with Edge Discovery APIs on AWS Wavelength

Achievements:

• Development of QoS profile for real-time, low-latency applications (QCI133)
• Jitter / Latency improvement by 20% using QCI133
• Patent submission, AI/ML to predict application performance give network conditions.

Dreamscape VR content evaluation over 5G

Dreamscape, an XR content provider for educational experiences, launched this initiative to assess the virtualization of NVIDIA GPU5 technology, exploring its functionality and limitations. The goal was to deploy the necessary infrastructure to onboard the application and deliver the product for evaluation at Arizona State University.

Achievements:

• Developed network slicing strategy for QCI133
• Full-Stack ecosystem deployment at scale, productivity improvement.
• Breakthrough technical evolution of XR content over Wavelength

AI/ML Network Performance Prediction

For this initiative, I developed a predictive model in Python using internal data sources to identify opportunities for enhancing coverage and throughput on Verizon’s network. The model helped prioritize capital investments to maximize improvements in customer experience.

Achievements:

• Developed AI/ML predictive models for network analysis
• Patent submission, AI/ML to recommend capital investment based on network improvement recommendations.
• This initiative led to a 25% reduction in network latency and a 20% increase in system throughput.

5G Technology Innovation

I developed an AI/ML model utilizing public data from FCC databases to identify potential antenna collocation opportunities in areas where tower companies were present. This initiative enabled me to create a five-year strategic plan for Verizon’s network expansion, optimizing time and capital investment while enhancing coverage and user experience.

Achievements:

• Developed AI/ML predictive models for network analysis
• Patent submission, AI/ML to identify colocation opportunities for network capacity grout.
• This initiative lead to a 30% improvement in network performance..