Lead ML & CV Engineer for ICM 2.0 Detection (Onsite)

Knightscope, Inc.; Sunnyvale (CA); Sur place; USD 200 000 - 240 000; Plein temps

Avantages offerts par ce poste

Medical

Dental

Vision

401(k)

Paid time off

Résumé du poste

Knightscope, Inc. seeks an experienced ML Engineer to own the ICM 2.0 detection models, leading the transfer from a third‑party vendor and driving ongoing improvements to precision, recall and latency.

You will manage training, evaluation, and deployment pipelines and collaborate with backend and MLOps teams to scale the solution onsite in Sunnyvale. The role requires deep learning expertise, production experience in audio/video detection, and a track record of owning model codebases from

Qualifications

• 7+ years of ML engineering with production CV experience.
• Audio/video detection model expertise.
• Strong deep learning framework skills (PyTorch preferred).
• Experience deploying models at scale.
• Comfort taking ownership of an existing model from a third party.

Responsabilités

• Own ICM 2.0 detection models and the training, evaluation, and deployment pipeline.
• Lead knowledge transfer from the current vendor on the detection engine.
• Partner with the video/streaming backend team on model serving and inference infrastructure.
• Partner with the MLOps Engineer on production model deployment and monitoring.
• Drive measurable improvement in detection precision, recall, and latency.

Connaissances

7+ years ML engineering

audio/video detection

deep learning (PyTorch)

productionizing models

ownership of model codebase

Outils

PyTorch

Description du poste

Knightscope, Inc. seeks an experienced ML Engineer to own the ICM 2.0 detection models, leading the transfer from a third‑party vendor and driving ongoing improvements to precision, recall and latency.

You will manage training, evaluation, and deployment pipelines and collaborate with backend and MLOps teams to scale the solution onsite in Sunnyvale. The role requires deep learning expertise, production experience in audio/video detection, and a track record of owning model codebases from