Drimva

Raydar

Machine Learning Engineer

$180K–$230K

San Francisco, California · Full-time · Posted Sep 30, 2026

Education
Bachelor's or Master's degree in computer science, machine learning or engineering, or equivalent proof through shipped production AI systems.

About the job

Build and deploy production multimodal machine-learning systems for a hardware product, from model orchestration and evaluation to constrained-device inference and interactive interfaces.

What you'll do

  • Build production image- and video-reasoning pipelines with multiple steps and tool calls; create evaluation and customer-feedback loops.
  • Optimize edge inference and runtimes, including model fallback for connectivity and latency constraints.
  • Train and fine-tune open-source models for on-premise deployment and develop interactive audio and visual interfaces.

What they ask for

  • Production experience delivering vision-language or other multimodal systems for real users, including models and orchestration.
  • Experience in an early-stage startup or another high-intensity engineering environment, with visible builder experience.
  • Bachelor's or Master's degree in computer science, machine learning, or engineering, or equivalent proof through shipped production AI systems.
  • Hands-on experience with fine-tuning, evaluation design, model orchestration, multi-step agent workflows using real tools, and deploying open-weight models on constrained or on-premise hardware, including quantization; able to work on-site five days a week.

Nice to have

  • Production experience with AR, wearable AI, or autonomous-driving computer vision.
  • Master's work involving vision or multimodal research; proficiency in Python and PyTorch.

Pay

USD 180,000 to 230,000 per year (base salary)

Tools

Python, PyTorch

AI focus

Production multimodal AI for a hardware product, including image and video reasoning, agent workflows, fine-tuning, evaluation, and optimized edge deployment.

Model-written summary · the original posting is authoritative

First seen Oct 30, 2026 · Last verified Nov 4, 2026 (UTC)