"quantization" Jobs

9 open tech roles matching “quantization”, taken straight from company career pages — not reposted from other job boards. Most in demand right now: Python, CUDA, GCP. Every listing is re-checked daily and closed roles are removed.

Showing 9 of 9 results

Inworld AI

Join Inworld AI as a Staff/Principal Machine Learning Engineer to optimize and serve state-of-the-art voice models in a fully remote role.

Inworld AI Switzerland Published 6 months ago
Flexible on stack
Reddit

Own performance optimization initiatives for production ML workloads in a fully remote role at Reddit.

Reddit Remote - United States $216.7k–$303.4k/yr Published 1 week ago
Flexible on stack
ElevenLabs

Join ElevenLabs as a Research Engineer to deploy and optimize AI models for real-time applications in a fully remote environment.

ElevenLabs United Kingdom Published 2 months ago
Flexible on stack
Horizon3.ai

Join Horizon3.ai as a Senior Machine Learning Engineer to operationalize AI models in a remote-first cybersecurity environment.

Horizon3.ai US, Remote $211k–$249k/yr Published 1 month ago
Flexible on stack
Inferact

Join Inferact as a staff engineer to work on optimizing AI inference across the vLLM stack in a fully remote role.

Inferact Remote Published 9 months ago
Flexible on stack
OpenRouter

Join OpenRouter as a Product Manager to build a marketplace connecting developers and inference providers in the AI ecosystem.

OpenRouter United States Published 2 days ago
Airbnb

Lead the fine-tuning and optimization of LLMs to enhance AI products at Airbnb with a focus on customer support.

Airbnb United States $292k–$365k/yr Published 5 months ago
Flexible on stack
OpenRouter

Join OpenRouter as the first AI Inference SRE to ensure operational health of provider supply in a fully remote role.

OpenRouter Remote (US) Published 2 weeks ago
Flexible on stack
OpenRouter

Lead a team in managing the provider ecosystem for AI routing at OpenRouter, focusing on technical leadership and hands-on coding.

OpenRouter Remote (US) Published 1 month ago
Flexible on stack Heavy meetings