Pulse
Machine Learning Engineer
$160K–$220K
San Francisco · On-site · Full-time · Posted Jul 30, 2025
- Experience
- 3+ years in applied ML or research, or a strong open-source record.
- Visa sponsorship
- Sponsorship available.
About the job
Develop, fine-tune, and productionize specialized vision and language models for document understanding and extraction.
What you'll do
- Train and fine-tune OCR, layout, table, and vision-language models; build evaluation, data-curation, and active-learning pipelines.
- Optimize GPU inference, batching, and quantization; productionize models with SLAs and rollback plans.
- Write internal notes to inform model and product roadmaps.
What they ask for
- 3+ years in applied ML or research, or a strong open-source record.
- PyTorch or JAX, and modern vision or multimodal architectures.
- Solid engineering discipline and metrics focus.
Nice to have
- Triton Inference Server, TensorRT, ONNX, or distributed training.
- Prior startup or founding experience.
Pay
$160K – $220K USD base salary
Tools
PyTorch, JAX, Triton Inference Server, TensorRT, ONNX
AI focus
Build and improve document-understanding models, including OCR, layout, table, and vision-language models, and related evaluation, data-curation, and active-learning pipelines.
Model-written summary · the original posting is authoritative