Embedded AI Engineer, On-Device Models
Core
Optimizing and deploying Deepgram's speech AI models for real-time, low-latency inference on resource-constrained embedded and edge devices.
Role type
Senior IC Embedded AI Engineer (On-Device Models)
Builds
Production-grade on-device voice agents and speech models for phones, wearables, and consumer hardware.
Domain
Consumer Electronics / Edge AI / Embedded Systems
Deliverable
production ML models
Required skills
C, C++, Rust, Model Quantization, Model Pruning, Knowledge Distillation, Operator Fusion, Bare-metal programming, RTOS development, NPU/DSP integration, Hardware-software co-design, Embedded Linux, Microcontroller development
Preferred skills
Real-time audio processing, DSP pipelines, Neural Architecture Search, Hardware benchmarking, Secure on-device deployment
Technologies
FreeRTOS, Zephyr, ONNX Runtime, TensorRT, TFLite, ExecuTorch
Responsibilities
Define architecture for on-device real-time inference across diverse processors; Optimize models for latency, memory, power, and thermal budgets; Write performance-critical runtime code; Integrate with edge inference runtimes and vendor toolchains; Build deployment pipelines and OTA update mechanisms; Establish benchmarking and validation frameworks; Partner with silicon vendors on SDK integration.
Seniority
Senior, hands-on IC
