Senior ML Performance Engineer

Senior ML Performance Engineer | $200-250K salary + equity | Early-stage ML Infrastructure. 30-50 employees, $50MM+ in funding

We’re seeking a seasoned ML Performance Optimization Specialist to spearhead the development and deployment of high-performance, scalable ML inference pipelines for a key early-stage company. You’ll optimize model performance, reduce latency, and maximize throughput for some of the most innovative companies in the world.

Key Responsibilities

• Model Optimization: Conduct in-depth performance profiling and analysis of ML models, identifying and eliminating bottlenecks.
• Pipeline Engineering: Design and implement efficient ML inference pipelines, leveraging technologies like TensorFlow Serving, TorchServe, and NVIDIA Triton Inference Server.
• Infrastructure Optimization: Collaborate with infrastructure teams to optimize hardware and software configurations for optimal ML performance, including GPU acceleration, distributed training, and model quantization.
• Performance Benchmarking: Develop and maintain rigorous performance benchmarks to measure and track improvements.
• Experimentation: Explore cutting‑edge techniques like model quantization, pruning, and knowledge distillation to further enhance performance.

Required Skills and Experience

• Strong proficiency in Python and ML frameworks (TensorFlow, PyTorch)
• Deep understanding of ML algorithms and architectures
• Expertise in performance optimization techniques (profiling, quantization, pruning, etc.)
• Hands‑on experience with container orchestration platforms (Kubernetes)
• Proficiency in cloud platforms (AWS, GCP, Azure)
• Strong problem‑solving and analytical skills

Preferred Qualifications

• Experience with ML hardware acceleration (GPUs, TPUs)
• Knowledge of distributed training frameworks (Horovod, DDP)
• Familiarity with MLIR and compiler optimization techniques

If you’re passionate about pushing the boundaries of ML performance and eager to work on cutting‑edge projects, we encourage you to apply.