Video Analytics Systems Engineer (SDE 3)

About Bytebeam
Bytebeam builds edge-to-cloud infrastructure for connected vehicles and devices. We are extending that foundation toward video-aware vehicle and fleet workflows where camera data, telemetry, diagnostics, and operational events can be understood together.

Role Overview

We are looking for a senior systems engineer to build production video-analytics infrastructure across the vehicle edge and cloud. You will own the software paths that ingest, encode, buffer, transport, index, and process video, then connect video events with vehicle telemetry. The work sits at the boundary of systems engineering, media pipelines, computer vision, and real-world network constraints.

What You'll Do

• Design edge and cloud pipelines for camera ingestion, encoding, buffering, upload, indexing, and event-triggered video capture.
• Synchronise video with GPS, CAN/OBD, IMU, alerts, and other vehicle telemetry.
• Integrate computer-vision models into resource-constrained edge and scalable cloud execution paths.
• Optimise latency, bandwidth, storage, compute, thermal load, and recovery under intermittent connectivity.
• Build observability for stream health, dropped frames, clock drift, model performance, and end-to-end delivery.
• Create replayable datasets and test harnesses that reproduce field conditions without exposing sensitive customer data.
• Partner with firmware, hardware, algorithms, backend, application, QA, and forward-deployed teams.
• Define secure handling, retention, access control, and auditability for video data.

Must-Have Skills

• 5+ years of software or systems engineering experience with production media, vision, edge, or high-throughput data systems.
• Strong proficiency in C++, Rust, Python.
• Hands-on experience with GStreamer, WebRTC or comparable media/vision tooling.
• Solid understanding of codecs, containers, timestamps, buffering, networking, concurrency, and performance profiling.

Good-to-Have

• Automotive camera, dashcam, ADAS, telematics, robotics, or surveillance-platform experience.
• Familiarity with PyTorch, ONNX, TensorRT, object storage, event-driven systems, or time-series databases.
• Knowledge of privacy-aware video retention, redaction, or secure device-to-cloud transport.

What Success Looks Like

Bytebeam has a measurable, production-ready video path that remains dependable in vehicles and poor networks, links events to the right telemetry and video window, and gives application teams stable building blocks for useful video-analytics workflows.