Job Description

Edge AI Lead Engineer - Austin, TX  

An exciting opportunity to join an early stage technology business developing physical AI and autonomous systems for complex, real world environments.

The business is building intelligent systems that combine advanced AI, computer vision and sensor technologies to help organisations understand activity, make faster decisions and operate autonomously in demanding environments.

With a growing engineering team and an ambitious product roadmap, they are looking for an experienced Edge AI Lead to take ownership of how their AI capabilities are deployed beyond the cloud and onto customer-controlled hardware.

Role

This is a hands-on technical leadership position focused on taking sophisticated perception, reasoning and AI models and making them work reliably on standalone, on-premise and embedded hardware.

You'll define the architecture, optimize models and runtimes, and solve the practical challenges of deploying AI within strict compute, memory, power and latency constraints.

From GPU servers and ruggedized workstations through to embedded platforms on vehicles and autonomous systems, you'll play a central role in determining how AI operates in disconnected environments where reliability and performance are essential.

Requirements

• Define the architecture for deploying perception, detection, tracking, re-identification and reasoning capabilities across edge and embedded hardware.

• Optimize and compress machine learning models using quantization, pruning, distillation and low-precision inference.

• Take large language models (LLMs), vision-language models (VLMs) and foundation models and make them viable on resource-constrained hardware.

• Develop efficient inference runtimes and services for standalone and air-gapped deployments.

• Profile and optimize GPU, CPU and NPU performance, balancing memory, thermal, power and latency requirements.

• Establish benchmarking frameworks and hardware-in-the-loop testing to validate performance and model accuracy.

• Work closely with AI, autonomy, systems and DevOps engineers to deliver production-ready edge deployments.

Experience Needed

• Proven experience deploying machine learning models into production on edge, embedded or otherwise resource-constrained hardware.

• Strong practical experience with model optimization techniques including quantization, pruning, distillation and low-precision inference.

• Deep understanding of GPU/CPU architecture, memory bandwidth, compute performance, power and thermal constraints.

• Experience with edge AI platforms and inference frameworks such as NVIDIA Jetson, TensorRT, ONNX Runtime or equivalent.

• Strong Python and C++, with the ability to work across model development, optimization and deployment.

• Experience with real-time computer vision workloads including object detection, tracking, segmentation and embeddings.

• Someone comfortable taking technical ownership, making architectural decisions and leading from the front in a small, senior engineering team.

Particularly interesting experience:

• LLM and vision-language model deployment on-device.

• Small language models, KV-cache optimization and speculative decoding.

• Multi-object tracking and re-identification.

• Air-gapped, disconnected or autonomous system deployments.

• Sensor integration across EO/IR, radar or RF.

• Experience in defense, public safety, government or other demanding operational environments.

This is an opportunity to join an ambitious early-stage business at an important stage of its technical development.  You'll have genuine ownership over the architecture and engineering foundations, work alongside experienced leaders across AI and autonomous systems, and help build technology that operates in the real world, not just R&D.

There is also scope to grow into broader technical leadership, including building and developing the edge engineering function as the business scales.

Candidates must be located, or willing to relocate to, Austin, TX "and must meet US Person requirements under ITAR.