LLM-Integrated Math Libraries Engineer

NVIDIA; Warszawa; On-site; PLN 221,000 - 507,000; Full time

Job summary

NVIDIA in Warszawa, Poland seeks a passionate engineer to accelerate API integration across the CUDA-X math library stack for modern LLM and agentic AI workflows. The role focuses on building high-quality, composable APIs and documentation that enable humans and AI agents to use GPU-powered libraries with ease.

You will collaborate with internal and external partners to deliver timely LLM-enhanced library releases, measure impact, and prototype tooling to translate APIs into AI-friendly

Qualifications

• PhD or MSc preferred (or equivalent experience) in a relevant field.
• Experience designing clear, composable APIs for complex domains.
• Strong written and verbal communication skills for collaboration.

Responsibilities

• Drive architecture and implementation of LLM-first APIs and docs.
• Collaborate with stakeholders to deliver timely LLM-enhanced releases.
• Build metrics and tooling to measure impact of AI-assisted code generation.
• Prototype tooling to convert APIs into AI-friendly representations.

Skills

LLMs

Finetuning

Reinforcement learning

RAGs

Agenting tooling

APIs design

Documentation

Education

PhD or MSc in Computational Science/CS/Applied Math

Tools

CUDA-X APIs (cuBLAS, cuFFT, cuSOLVER, cuSPARSE)

Job description

NVIDIA in Warszawa, Poland seeks a passionate engineer to accelerate API integration across the CUDA-X math library stack for modern LLM and agentic AI workflows. The role focuses on building high-quality, composable APIs and documentation that enable humans and AI agents to use GPU-powered libraries with ease.

You will collaborate with internal and external partners to deliver timely LLM-enhanced library releases, measure impact, and prototype tooling to translate APIs into AI-friendly