Lead Software Engineer, Model Serving Platform

What this role involves

Sciforium is seeking a Lead Software Engineer to architect and lead the development of their model serving platform. The role involves building high-performance C++ and CUDA/HIP modules, collaborating with ML researchers, and mentoring engineers. The position requires 5+ years of experience and is onsite.

Skills this posting asks for

• c++
• python
• cuda
• hip
• kubernetes
• ray
• containerization
• llm inference
• batching
• kv cache
• distributed systems
• gpu kernels
• performance profiling
• observability
• vllm
• sglang
• trt-llm
• mlops
• rocm
• multimodal models

Requirements

• 5 years of experience
• Level: senior
• Remote policy: onsite

From the employer’s posting

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 ful…