At Neurobus, we develop intelligent vision systems powered by neuromorphic technologies to enable ultra-energy-efficient embedded intelligence, allowing autonomous systems to perceive, navigate, and respond in real time across drones, defense, and space.
Neuromorphic technology isn't "plug-and-play." Deploying Spiking Neural Networks (SNNs) and event-based vision onto heterogeneous processors requires more than just calling an API.
We are looking for an Edge AI Engineer who thrives at the intersection of hardware and software. Your mission is to architect and optimize high-performance inference pipelines that handle multi-stage AI models, novel sensor data (Event-based cameras, LiDAR), and constrained compute environments where every microsecond and microjoule counts.
As a core contributor to our edge AI pipelines, you will:
Architect Inference Pipelines: Design and implement end-to-end data processing pipelines (from raw sensor ingest to actuation) optimized for ultra-low latency and minimal power footprints.
Hardware-Software Co-Design: Evaluate and benchmark next-gen hardware (NPUs, FPGAs, Neuromorphic ASICs). You will provide the feedback loop that shapes our future hardware specifications.
Optimize for the Metal: Move beyond high-level Python wrappers. You will profile, prune, and quantize models, writing high-efficiency C++ code to extract maximum performance from heterogeneous compute stack (CPU/GPU/NPU).
Build the Production Engine: Establish "Edge-Ops" best practices. You will lead the way in CI/CD for embedded systems, automated hardware-in-the-loop (HIL) testing, and versioning for model-on-chip deployments.
Collaborate on Intelligence: Work alongside AI Researchers to translate theoretical models into production-ready software that meets the stringent reliability standards of the Space and Defense sectors.
Required:
Educational background: Master’s or PhD in Computer Science, Robotics, Embedded Systems, or a related technical field.
Edge AI experience: 3+ years of experience optimizing AI pipelines for edge platforms. You are fluent in Python and Modern C++ (14/17/20), with a deep understanding of memory management and asynchronicity.
Computer Vision Depth: Previous experience with computer vision libraries (OpenCV/Numpy) and deep learning frameworks (PyTorch/TensorFlow). You understand the internals of deep neural networks.
Edge deployment expertise: Hands-on experience with edge inference engines (e.g., TensorRT, OpenVINO, ONNX Runtime), deploying to constrained SoCs (Nvidia Jetson, NXP i.MX, Axelera, STM32, Hailo, or similar) and interfacing with dedicated NPU accelerators.
Systems fluency: Comfortable working in Linux environments, understanding RTOS concepts, configuring edge platforms and their OS, and managing containerized deployments (Docker) on the edge.
Strong communication and collaboration skills, with a proactive approach to teamwork.
Fluency in English (written and spoken).
Bonus points:
Experience with event-based vision (Prophesee/DVS) or neuromorphic frameworks (Lava, Tonic, etc.).
Experience with GStreamer for high-throughput video handling.
Familiarity with ROS2 or other robotics middleware.
Processor-specific optimization skills (e.g. assembly, SIMD, CUDA / OpenCL, etc.).
A track record of contributions to open-source software.
Three rounds of interview, with an online coding test after the first one.
