HN Who's Hiring💵 paid workby snzkescored by google/gemini-2.5-flash

Laminar Engineering | Full-Stack CV/ML Systems Engineer | Contract | REMOTE Laminar Engineering builds real-time compute

Visit project ↗Discussion ↗

💵 Funded work— see “Deliver it” below for how to respond.

Opportunity

AI-buildable

Traction

Creativity

The take

effort: ~3+ months

This is a contract role for a Full-Stack CV/ML Systems Engineer at Laminar Engineering, focused on building real-time computer vision systems for airborne drone detection and tracking. The role spans the entire pipeline, from data collection and model training to NVIDIA edge deployment, real-time video processing, and operator dashboards. It requires deep expertise in computer vision, ML, and embedded systems.

Demand & gap

Painkillerdemand 89

Demand

Will. to pay

Gap

Buyer

Enterprise / orgs

The gap in what exists: There is a strong demand for highly specialized engineers who can deliver end-to-end real-time CV/ML systems for critical applications, a niche not easily filled by generalist teams or off-the-shelf solutions.
Wedge to win: Position oneself as a highly specialized solo contractor capable of rapidly prototyping and delivering specific, high-impact components of the overall system, leveraging AI coding tools for speed.
Reputation value (worth doing free for proof) — 100/100: Successfully delivering on a contract for real-time airborne drone detection provides exceptional proof-of-work and builds strong credibility in high-stakes CV/ML and embedded systems.
Likely monetization: high-value contract / project-based services
Incumbents to beat: Specialized CV/ML consulting firmsDefense contractors with in-house R&D

Deliver it

A starter prompt for Claude Code, what you'll need, and how to reach them.

You are a Full-Stack CV/ML Systems Engineer for Laminar Engineering. Your task is to develop a real-time computer vision module for airborne drone detection and tracking on NVIDIA edge devices (Jetson). Focus on the core components for an MVP: video stream input, a lightweight object detection model (e.g., YOLOv8-nano), multi-object tracking, and a basic API for results. Utilize Python, PyTorch, TensorRT for inference optimization, and DeepStream/GStreamer for the video pipeline. Assume a simulated video feed of drones as input. The output should be a Dockerized application runnable on a Jetson, providing bounding box coordinates and track IDs for detected drones. 

**MVP Slice:** Implement a proof-of-concept for real-time drone detection and tracking on a pre-recorded video file using a lightweight PyTorch model converted to TensorRT. 

**Stack:** Python, PyTorch, TensorRT, DeepStream/GStreamer, Docker. 

**Build/Verify Gate:** The system successfully processes a sample video file, detects drones, tracks them, and outputs their positions and IDs in real-time, displaying this information in a terminal or a simple web interface. The solution should be packaged as a Docker image for easy deployment.

Prerequisites — cost & what to learn

How you'd build it

1Deep dive into NVIDIA Jetson, TensorRT, and DeepStream documentation and tutorials for edge CV deployment.
2Set up a local development environment with PyTorch, OpenCV, and relevant NVIDIA SDKs.
3Develop a minimal viable product (MVP) for object detection and tracking on simulated video feeds using a pre-trained model.
4Integrate a basic frontend dashboard to visualize the processed video and detection results.
5Gain practical experience with embedded Linux and camera/gimbal control, potentially with open-source hardware.
6Build a robust CI/CD pipeline for model deployment and system updates to edge devices.

Risks & moats

The required expertise across CV/ML, real-time systems, embedded Linux, and hardware is vast for a solo operator.
Proprietary data collection and client-specific requirements may be difficult to replicate or access.
Diagnosing and fixing real-world failures in live airborne systems requires significant domain experience and physical access.
The 'clear end state' is still vague, implying a need for proactive definition and potential scope creep in a solo capacity.

Original context

Laminar Engineering | Full-Stack CV/ML Systems Engineer | Contract | REMOTE Laminar Engineering builds real-time computer vision systems for airborne detection and tracking of drones, combining multispectral sensing and edge inference. We’re delivering this for a client, and we’re looking for a versatile engineer to join the R&D effort. This is a role for someone who thrives with autonomy when given a clear end state. More details to be shared with the right candidates. The role spans the whole pipeline, from data to deployed system: CV/ML on NVIDIA edge devices (Jetson / TensorRT / DeepStream), real-time video processing and multi-object tracking, camera/gimbal control, dataset collection and curation, model training and evaluation, backend infrastructure and APIs, and operator-facing dashboards. You’d own deployed vision systems end to end, including diagnosing and fixing real-world failures in live video. Strong fit: applied CV and perception, including tracking, real-time pipelines, and NVIDIA edge deployment, plus strong Python and comfort across the stack. Relevant experience: PyTorch, TensorRT/ONNX, DeepStream/GStreamer, embedded Linux, and frontend work for operator tooling