Research Engineer - On-device

About the role

This stealth startup is developing an AI-native platform designed to automate IT workflows by resolving technical issues directly on end-user devices. You will focus on deploying machine learning models that operate locally on Windows, Mac, and Linux systems to replace manual ticket management and reactive support.

Key facts

What you'll do

• Build and optimize lightweight machine learning models for local execution on desktop operating systems.
• Manage inference performance within strict memory and compute limitations.
• Design pipelines that span from initial model training and evaluation to final deployment.
• Integrate on-device intelligence with backend reasoning systems and agent-based workflows.
• Process high-volume telemetry and behavioral data to ensure reliable autonomous diagnostics.

Requirements

• Proven experience deploying machine learning models in resource-constrained environments such as mobile, embedded, or edge hardware.
• Practical history of moving models from research prototypes into live production systems.
• Ability to thrive in an early-stage company environment with high levels of individual responsibility.
• Expertise in edge ML optimization and real-time inference techniques.

Nice to have

• Background in developing multimodal systems incorporating vision, audio, or sensor inputs.
• Experience building agent-based frameworks or LLM-driven automated workflows.
• Track record of bridging the gap between experimental research and functional software products.

Skills & tools

• Edge ML
• On-device inference
• Model optimization
• LLM and agent systems
• Multimodal data processing

Practical notes

The company is led by a serial founder with a history of successful exits. This is a small team environment offering significant influence over the product roadmap.