Computer Vision Engineer — On-Device Detection (Founding Contractor). Remote (US time zones preferred) | Contract | Pre-revenue, deferred cash + equity

ABOUT US. BullzEye is a pre-revenue golf technology company building something the golf industry has not seen. Our iPhone app scans a putting green and reads the surface, entirely on device. No cloud round trip. The core problem is real-time surface reconstruction and object detection on consumer hardware, with no server in the loop. Golf is the first application, not the last one. Patent stage, public beta November eighth. The detector is the heart of the product, and it is not good enough yet. That is the job.

WHAT YOU’LL OWN. Ball and cup detection end to end, from a working but mediocre on-device baseline to shipping quality. Find where it fails, small objects, hard lighting, wet greens, long putts, clutter, and drive measurable improvement. Make the training pipeline reproducible. Own on-device performance: export, quantization, non-max suppression baked in, benchmarked to run inside a real-time scan budget on a phone. Own calibrated refusal, because the product would rather decline a bad frame than give a wrong read. Own the evaluation harness: per-condition slicing, held-out scoring, regression tracking across versions. We ship a closed-source binary, so training stays on permissively licensed stacks.

WHAT WE’RE LOOKING FOR. You have shipped production object detection in a real domain. Not notebooks, not a fine-tune you never deployed. You have run inference on device or at the edge and you can talk about what you gave up to get there. You have made a model small and fast, not just accurate. You are fluent in real evaluation, mAP, PR curves, per-slice analysis, domain shift, hard negatives. You scope down to the simplest thing that measurably works. We are ten weeks from a hard date, so cutting scope is not a soft skill here.

SENSING BACKGROUND IS OPEN. The depth source today is the iPhone’s built-in sensor, but the architecture is source agnostic. Stereo, structured light, time of flight, LiDAR, and near-infrared backgrounds all map. Having shipped iPhone LiDAR specifically is not required.

LOGISTICS. Contract to start, remote, US time zones. You must be authorized to work in the US now. We cannot sponsor, and we cannot wait on pending work authorization for this role. Pre-revenue, so cash is deferred and the real compensation is equity, zero point two five to zero point five percent. The salary field shows a platform minimum, not the offer. Wellfound requires a cofounder job type for equity-weighted roles, so read this as a founding contractor engagement. We are closing a bridge round now and expect to raise a seed in late fall. The intent is to convert this to a paid contract once financing closes, on terms agreed at that point. That is intent, not a guarantee. Take this role on the equity, not on the expectation of cash.

HOW TO APPLY. Email [email protected] directly.

Subject line: Detection. Tell me about one detection model you took from promising to actually shipped. What device or runtime it ran on, what frame rate you hit, and what was hard about the last ten percent. Three paragraphs is plenty. Applications that skip this question do not get read.