Software Engineer, Inference
Hybrid (San Francisco, CA)$150k - $230kFull-timePosted 11mo ago
About the role
Pulse is building production-grade, unstructured document extraction. The role focuses on low-latency, high-throughput inference for OCR and multimodal models, with a multi-stage architecture involving layout understanding, fast OCR, reading-order algorithms, table parsing, and vision-language models. This position is based in San Francisco and requires 3+ years of experience, US visa/sponsorship available, and in-office presence.
Responsibilities
- Build inference services with smart batching and caching
- Optimize kernels, tokenization, and model graphs
- Evaluate vLLM, TensorRT LLM, and Triton tradeoffs
- Implement autoscaling and admission control with clear SLOs
- Own performance dashboards and capacity planning
Qualifications
- 3+ years in performance engineering or ML systems
- Strong Python, plus C++ or CUDA exposure
- Experience with GPU profiling and model serving
Skills
Required:PythonC++ or CUDAGPU profilingModel serving
Benefits
- Relocation assistance for Bay Area moves
- Daily meal stipend
- Comprehensive medical, vision, and dental coverage
- Equity and performance-based bonuses
Last updated Sep 16, 2026. Interested? Apply on the company site.
About Pulse
Production-grade unstructured document extraction
Pulse converts complex information into LLM-ready inputs. Our API supports all document formats, from PDFs to Word, Excel, etc. Pulse integrates seamlessly with any existing data pipeline in minutes without any training or complexity.
More open roles at Pulse
- Growth and Marketing1mo agoGrowth and Marketing
- Business Development Representative (BDR)10mo agoHybrid (San Francisco) · Sales
- Machine Learning Engineer11mo agoHybrid (San Francisco) · Engineering
- Design Engineer11mo agoSan Francisco
- DevOps Engineer11mo agoHybrid (San Francisco)