Backend Engineer, Video Analytics and Predictive

Аналитика рынка

Стек

Бесплатно. Вход по почте.

Откликнуться

Загрузите резюме — мы свяжем вас с работодателем напрямую через нашу базу.

Описание

We build sealed AI appliances for factories. AJetson-class box on the plant network watches the cameras. A DGX Spark classbox with 128 GB of unified memory holds the record of what happened and readsthe tool sensors. Only text crosses between the two, and video is fetched ondemand with an audit record of every viewing. You will build the pipeline that turns camera streamsinto typed events and tool sensor streams into early alerts that name theaffected sensors. The first release is classical computer vision and timeseries. The language model layer comes after that works. WHAT YOU WILL BUILD Edge detection and tracking. RTSP from the DVRs into GStreamer or DeepStream, inference through ONNX Runtime or TensorRT,and a frame budget per camera that holds when all thirty cameras are busy. Typed events. One event per track with a stable identifier, a confidence, a one-sentence caption and a pointer to the clip, published over MQTT with a local spool that drains after a network cut. Central ingest. PostgreSQL with TimescaleDB. Deduplicate, place the event on the site plan through the camera calibration, correlate into incidents, aggregate so every dashboard panel answers under a second. Equipment Watch. Multivariate sensor streams through RFOD and CAD, so an alert carries the affected sensors and a feature attribution rather than a bare score, measured against labelled failure windows. Replay. Role check, audit record, signed ticket, WebRTC relay from the edge box to the browser, held

Полное описание, компания и ссылка на вакансию — после входа. Войти

Контакты работодателя (email/phone/telegram) скрыты из публичного превью — отправьте резюме, чтобы мы связали вас напрямую.