Embedded AI Engineer

HarkSan Jose Embedded Software

Apply on company site

Opens the official Hark careers page

Salary

$200K - $450K USD

Job type

Full-time

Posted

2 Sept 2026

Status

Open · checked 7d ago

About this role

The Embedded AI Engineer position at Hark in San Jose is an exciting opportunity for an experienced professional to develop the full on-device AI stack for next-generation hardware. This role suits individuals with a strong background in machine learning and embedded systems who want to build proactive, multimodal intelligence.

Role Overview

In this capacity, the selected professional will collaborate directly with the AI research team to integrate artificial intelligence into Hark's cutting-edge hardware devices. The position involves overseeing everything from data ingestion and model architecture co-design to optimization and deployment on embedded hardware. The ideal candidate understands the performance limits of edge AI models and embedded system constraints regarding compute, memory, and power.

Key Responsibilities

Build scalable data collection and ingestion pipelines for embedded systems that utilize various sensors.

Partner with model teams to co-design architectures that satisfy strict latency, memory, power, and bandwidth targets.

Coordinate with platform vendors to deploy toolchains, SDKs, and new hardware accelerators for optimized model execution.

Integrate machine learning inference directly into embedded firmware using C, C++, or Rust.

Profile and fine-tune real-time performance, memory utilization, and power usage.

Assess and choose appropriate silicon platforms, such as GPUs and NPUs, for edge and on-device deployment.

Requirements

Five years of total experience in machine learning engineering.

At least two years of focused experience working with embedded or edge machine learning.

Strong familiarity with embedded systems and CPU, DSP, or NPU hardware architectures.

Hands-on background utilizing sensors like microphones, gyroscopes, accelerometers, and IMUs.

Experience in constructing data collection pipelines for sensors.

Knowledge of embedded machine learning runtimes such as llamacpp, TFLite, and QNN.

Proven skill in optimizing machine learning models for edge processors and microcontrollers like RISC-V, ARM Cortex-M/A, and DSPs.

Practical experience deploying workloads onto NPUs or specialized embedded accelerators.

Nice to Have

Background working with voice, audio, or vision models.

Familiarity with lightweight large language models.

Knowledge of edge AI model performance traits, including transformers, CNNs, RNNs, KV-cache behavior, and memory bandwidth demands.

Experience building hybrid edge-LLM pipelines or integrating small language models directly on devices.

Previous work on Internet of Things, wearables, industrial sensing, or robotics products.

Skills and Tools

Programming Languages: C, C++, Rust

Runtimes & Frameworks: TFLite, QNN, llamacpp

Hardware Architectures: ARM Cortex-M/A, RISC-V, GPUs, NPUs, DSPs

Sensors: IMUs, microphones, gyroscopes, accelerometers

Salary and Benefits

The US base salary range for this full-time position spans from $200,000 to $450,000 per year. Actual pay may vary depending on individual factors such as relevant skills, job knowledge, and experience. Additional components or benefits may also be included in the total compensation package.

About the Company

Hark is an artificial intelligence company creating advanced, personalized, and multimodal intelligence that interacts with the real world through text, speech, vision, and persistent memory. The company pairs this intelligence with next-generation hardware to form a unified interface for agentic systems.

How to Apply

Interested candidates should visit the official company careers page to submit their application for this position.

Skills and tech stack

Rust C++LLM

Questions candidates ask

What is the location of the Embedded AI Engineer role?

The job is located in San Jose.

What is the salary range for this position?

The US base salary ranges from $200,000 to $450,000 annually.

How many years of experience are required?

The position requires 5 years of experience in machine learning engineering, with at least 2 years focused on edge or embedded ML.

What programming languages are used for embedded firmware integration?

Embedded firmware is written in C, C++, or Rust.

How should I apply for this job?

Candidates should apply directly on the official company careers page.

Ready to apply?

The application is on the official Hark careers page.

Apply on company site