Distributed Training and Inference Engineer
US$190,000 – US$250,000 / yr base salary
Posted 24 Aug 2026
BranchFactor Summary
The original posting with the fluff stripped out
- Management level
- Individual contributor
- Employment
- On-Site
- Contract
- Full-time
| Base salary | US$190,000 – US$250,000 / yr |
|---|
Field of studyComputer Engineering, Computer Science, Electrical Engineering (or related)
Study levelBachelor's or higher
Sciforium is an AI infrastructure company developing next-generation multimodal AI models and a proprietary, high-efficiency serving platform. Backed by multi-million-dollar funding and direct sponsorship from AMD with hands-on support from AMD engineers the team is scaling rapidly to build the full stack powering frontier AI models and real-time applications.
About the role
Sciforium is seeking a highly skilled Distributed Training and Inference Engineer to build, optimize, and maintain the critical software stack that powers our large-scale AI training and serving workloads. In this role, you will work across the entire machine learning infrastructure from low-level CUDA/ROCm runtimes to high-level frameworks like JAX and PyTorch to ensure our distributed training systems are fast, scalable, stable, and efficient.
This position is ideal for someone who loves deep systems engineering, debugging complex hardware–software interactions, and optimizing performance at every layer of the ML stack. You will play a pivotal role in enabling the training and deployment of next-generation LLMs and generative AI models.
What you'll do
-
Software Stack Maintenance: Maintain, update, and optimize critical ML libraries and frameworks including JAX, PyTorch, CUDA, and ROCm across multiple environments and hardware configurations.
-
End-to-End Stack Ownership: Build, maintain, and continuously improve the entire ML software stack from ROCm/CUDA drivers to high-level JAX/PyTorch tooling.
-
Distributed System Optimization: Ensure all model implementations are efficiently sharded, partitioned, and configured for large-scale distributed training and serving.
-
System Integration: Continuously integrate and validate modules for runtime correctness, memory efficiency, and scalability across multi-node GPU/accelerator clusters.
-
Profiling & Performance Analysis: Conduct detailed profiling of compilation graphs, training workloads, and runtime execution to optimize performance and eliminate bottlenecks.
-
Debugging & Reliability: Troubleshoot complex hardware–software interaction issues, including vLLM compilation failures on ROCm, CUDA memory leaks, distributed runtime failures, and kernel-level inconsistencies.
-
Collaborate with research, infrastructure, and kernel engineering teams to improve system throughput, stability, and developer experience.