xelys jobs xelys jobs

Jr Software Engineer - ML Runtime - Rust

tensordyne

juniorbackendengineering-management Munich 100 days ago via Arbeitnow

See how well this job matches your profile

Sign up to get an AI match score and generate a tailored application in seconds.

Get your match score

Tags

RustML InferenceLow-Level SystemsEmbedded SystemsLLMsAttention ArchitecturesPerformance OptimizationDebuggingGoogle Cloud Platform (GCP)Python

About the role

Role overview

You will join the Low-Level Runtime team to build components of an ML model execution framework. The work focuses on low-level, assembler-style architecture hyper-optimized for Tensordyne hardware, as well as porting ML model components to run on that runtime.

Responsibilities

  • Maintain and develop a low-level ML model execution framework in Rust
  • Maintain and enhance components and full models targeting the company’s hardware
  • Optimize ML model performance to maximize throughput and utilization
  • Debug and productize models across virtual and physical hardware environments

Preferred qualifications

  • Experience with Rust
  • General awareness of ML model concepts, including LLMs and attention architectures
  • Experience with low-level development (e.g., embedded systems)
  • Experience with driving agents and building LLM engineering workflows

Nice to have

  • Experience with Google Cloud Platform (GCP) or similar
  • Experience with Python
  • Deep knowledge of modern LLM attention architectures (e.g., GDN, DSA, KDA)

About tensordyne

Tensordyne (formerly Recogni) is an AI system solutions company building very high-performance, low-power generative AI inference systems. It combines custom silicon with hardware and software to accelerate multimodal generative AI inference at scale for hyperscalers and neocloud data center customers. The company is headquartered in Sunnyvale (CA) and Munich (Germany).

Scraped 8/13/2026