Summer 2027 PhD AI Research Infrastructure, RL Post-Training Intern - CA, Santa Clara, United States
Originally posted 21 September 2026 by the employer.
Develop and optimize infrastructure for RL-based post-training of large language and multimodal models.
About the role
This internship involves developing and optimizing AI research infrastructure for RL-based post-training of large language and multimodal models.
What you'll do
- Develop and optimize infrastructure for RL-based post-training of large language and multimodal models.
- Build scalable systems for rollout generation, inference, reward computation, and policy updates.
- Improve distributed training efficiency, reliability, fault tolerance, and resource utilization.
- Design interfaces that enable researchers to implement and evaluate new RL algorithms quickly.
- Build tools for experiment configuration, checkpointing, logging, monitoring, and reproducibility.
- Profile end-to-end training pipelines and resolve performance, memory, and communication bottlenecks.
- Support on-policy and off-policy training workflows using verifiable, preference-based, or model-generated feedback.
- Collaborate with researchers to translate experimental requirements into production-quality infrastructure.
- Document system designs and contribute to technical reports and publications.
What you'll need
- Currently pursuing a PhD in Computer Science, Machine Learning, Artificial Intelligence, Computer Engineering, or a related field.
- Strong programming skills in Python and experience with PyTorch.
- Knowledge of reinforcement learning, LLM post-training, RLHF/RLAIF, or preference optimization.
- Experience with distributed training, multi-GPU workloads, or large-scale inference.
- Familiarity with parallelism strategies such as data, tensor, pipeline, or sequence parallelism.
- Experience with training and inference frameworks, orchestration systems, or cluster environments.
- Understanding of GPU performance, memory management, networking, and distributed communication.
- Experience building reliable research infrastructure, profiling systems, or debugging distributed workloads.
Nice to have
- Familiarity with containerization, experiment tracking, and cloud or cluster computing.
- Publications at venues such as ICML, NeurIPS, ICLR, MLSys, CVPR, ICCV, or ECCV.
Skills: AI Research Infrastructure, RL-based post-training, scalable systems, distributed training, multi-GPU workloads, PyTorch
This role has been open 20 days — well below the 53-day median for Infrastructure/Platform Software roles.
Infrastructure/Platform Software · Infrastructure Platform
|
Open roles in category
923
|
Median days open
53 d
|
Median salary
$229k
|
See the full market breakdown ▾Category comparison, skills in demand, and who else is hiring
| Metric | AMD | All employers we track in this specialty (923 roles · 86 employers) |
|---|---|---|
| Open roles in this specialty | 51 | 923 |
| Open roles in the wider Software, Firmware & Systems family | 407 | 6028 · 146 employers |
| Median days open | 31 d | 53 d (−22 d vs this employer) |
| Median salary (USD postings) | — | $229k |
Skills observed across this category: AI Research Infrastructure, RL-based post-training, scalable systems, distributed training, multi-GPU workloads, PyTorch
Who's hiring in this category
- NVIDIA · 255 open roles · median 48 d
- Qualcomm · 60 open roles · median 60 d
- AMD (this employer) · 51 open roles · median 30 d
- Graphcore · 44 open roles · median 86 d
- Cerebras · 39 open roles · median 76 d
- Intel Corporation · 36 open roles · median 15 d
How we counted: 923 open Infrastructure/Platform Software (Infrastructure Platform) roles from 86 employers tracked in the SemiconductorJobs index, counted 11 Oct 2026. Specialty figures count only roles carrying this exact specialty label, so an employer's related work in neighbouring specialties is not included there — it is counted in the wider Software, Firmware & Systems family row. Figures refresh nightly.