GenAI Intern

Descripción del puesto:

Number of Position(s): 2
Duration: 10 Weeks
Date: June - August, 2026
Location: On-site, in Murray Hill, New Jersey.

Education Recommendations

Eligible candidates should currently be pursuing a Master's or Ph.D. in Computer Science, Computer Engineering, or a related field with an accredited school in the US.

The selected candidate will have the opportunity to contribute to a Machine Learning Operations (MLOps) platform, which supports state-of-the-art training and inference features with a focus on sustainable MLOps practices.
The selected candidate will have the opportunity to contribute to a GenAI and AI/ML systems. The selected candidates will work on:

* Help design and prototype AI-driven inference mechanisms for next-generation radio access networks (RAN).
* Classifying diverse AI/RAN workloads (e.g., CSI/beamforming models, traffic prediction, anomaly detection, RAN policy/LLM agents).
* Defining multi-dimensional SLA requirements for each, and exploring how these workloads can be dynamically placed, migrated, or optimized across heterogeneous compute tiers including DU/CU, edge, and cloud GPU nodes.
* Telemetry extraction from RAN and compute layers, workload-specific SLA modeling, and building small-scale simulations or testbeds that demonstrate intelligent inference routing and performance adaptation under real-world constraints such as latency, jitter, accuracy, and compute saturation

Requerimientos del candidato/a:

We are looking for students with the following background and skillset.
* Strong knowledge of RAN concepts: DU/CU functional split, CSI, beamforming, HARQ timing, scheduling pipelines
* Machine learning inference experience: PyTorch, ONNX, Hugging Face
* Familiarity with AI models used in RAN: CSI compression networks, digital twin models, traffic prediction, or RAN automation agents
* Python proficiency for research prototyping
* Experience with real-time or near-real-time systems (latency-sensitive pipelines)
* Strong knowledge of latency/jitter, QoS, and SLA concepts
* Familiarity with distributed systems or network simulation tools (Mininet, ns-3, OAI, srsRAN)