Principal AI Researcher

Cisco•Milpitas, CA

5d•$296,300 - $431,000

About The Position

The AI Software & Platform Group builds Cisco's enterprise AI platforms and models, bringing together AI with Cisco's networking, security, and observability portfolio. Our research team is working on a fundamental problem: Can we build models that understand how networks behave well enough to operate them autonomously? We are developing world models of networks that learn from topology, telemetry, configuration, traffic, events, and operational history. The goal is to build models that can understand the current state of a network, predict what happens next, simulate the impact of an action, and provide the foundation for self-driving networks.

Requirements

Masters in Computer Science, Machine Learning, Artificial Intelligence, Robotics, Autonomous Systems, or a related field, with 8+ years of relevant research experience, or equivalent demonstrated research impact.
Research experience in one or more of world models, reinforcement learning, model-based RL, representation learning, planning, or sequential decision-making.
3+ years of hands-on experience building, training, and optimizing deep learning models using PyTorch, JAX, or comparable frameworks, including at least 2 models trained and evaluated at production or research scale.
Demonstrated track record of research through publications at leading AI/ML conferences or equivalent research contributions.
Ability to take a research problem from an idea through experimentation and working prototype.

Nice To Haves

Experience building world models, learned simulators, or model-based RL systems.
Experience with offline or online reinforcement learning, simulation, or learning from large-scale trajectories.
Experience with graph learning or modeling complex dynamic systems.
Experience in networking, distributed systems, robotics, autonomous systems, or embodied AI.
Experience applying research to real-world systems at scale.

Responsibilities

Lead research on world models and learning-based approaches for autonomous networks.
Develop world models that learn the state and dynamics of large-scale network environments.
Build models that predict network behavior and the impact of configuration and operational changes.
Develop model-based reinforcement learning approaches for network optimization, troubleshooting, and remediation.
Build simulation environments for training and evaluating models and agents before they operate on production networks.
Explore new approaches to representation learning, planning, and sequential decision-making for network environments.
Develop methods to evaluate the reliability and safety of learned policies before deployment.
Work with networking experts and engineers to evaluate research on real network data and systems.
Publish research at leading AI/ML conferences and contribute to open-source projects.