Remote On-Device AI Engineer: Embedded Models

Bazeta; United States; Remote; USD 120,000 - 160,000; Full time

Job summary

Deepgram is seeking an Embedded AI Engineer to optimize and deploy speech models on resource-constrained hardware. You will write kernels and operators for embedded SoCs, DSPs, and NPUs, and collaborate with ML teams to adapt models to edge devices.

This role focuses on low-level performance and portable runtimes across embedded Linux, bare-metal, and RTOS environments. Ideal candidates have strong C/C++/Rust skills, hands-on experience with on-device optimization, and familiarity with edge

Qualifications

• Strong proficiency in C, C++, and/or Rust for performance-critical code.
• Experience delivering production systems on resource-constrained hardware.
• Hands-on experience with model optimization for on-device deployment.
• Familiarity with edge inference runtimes (ONNX Runtime, TensorRT, TFLite, ExecuTorch) and vendor toolchains.

Responsibilities

• Write and optimize custom kernels and operators (C, C++, Rust) for embedded platforms and NPUs.
• Own target-side optimization: quantization, fusion, memory layout, and platform-specific compilation.
• Integrate with vendor NPU/DSP toolchains and edge runtimes; extend where needed.
• Deliver kernels and runtimes as reusable building blocks for ML engineers.
• Build runtime code for embedded environments (embedded Linux, bare-metal, RTOS).
• Establish per-platform benchmarking and validation for latency, power, and memory.
• Collaborate with silicon vendors on SDK integration and low-level tuning.
• Provide feedback to ML and Research teams to ease landing models on constrained targets.

Skills

C

C++

Rust

Model optimization

Edge AI

Tools

ONNX Runtime

TensorRT

TFLite

ExecuTorch

Job description

Deepgram is seeking an Embedded AI Engineer to optimize and deploy speech models on resource-constrained hardware. You will write kernels and operators for embedded SoCs, DSPs, and NPUs, and collaborate with ML teams to adapt models to edge devices.

This role focuses on low-level performance and portable runtimes across embedded Linux, bare-metal, and RTOS environments. Ideal candidates have strong C/C++/Rust skills, hands-on experience with on-device optimization, and familiarity with edge