Ship GenAI research into production ML pipelines at Google DeepMind

This Google DeepMind Software Engineer role (Sunnyvale, hybrid) sits at the boundary of research and production: prototyping GenAI solutions for generative media and multimodal understanding, building ML pipelines, and hardening product code with integration/performance/security testing. It requires 2 years of experience training generative AI models for media generation, building models in TensorFlow/PyTorch/JAX, managing ML infrastructure (deployment, evaluation, optimization, data processing), and general software development in Java/C/C++/Python/Go. Expect the mock to blend ML system design with core coding fundamentals (data structures/algorithms) and production-engineering rigor.

走进这场面试

免费 · 一场按这个岗位校准的真语音模拟

练这场面试

这场面试考什么

Training and evaluating generative AI models for media generation
ML framework tradeoffs (TensorFlow, PyTorch, JAX)
ML infrastructure: deployment, evaluation, optimization, data pipelines
Core data structures and algorithms
Debugging and root-causing production system issues (performance, security, reliability)