Video Platform Engineer - E3

About this job

Instawork is hiring a Video Platform Engineer - E3 for its Bengaluru team. Instawork operates an AI-powered labor marketplace that connects skilled hourly professionals with local businesses, and through Instawork Robotics Labs (IRL), it captures high-fidelity task data from commercial and residential environments to help robotics labs train foundation models. In this role, you will own the pipeline that transforms raw video capture from multi-camera rigs into training-ready data. Your responsibilities will combine media-engineering and distributed systems, focusing on ingest, transcoding, time-synchronization, automated quality checks, indexing, and delivery. On a normal day, you might design resilient cloud storage ingestion paths, make codec and container decisions, build automated quality checks for video streams, optimize storage and GPU processing costs, and collaborate directly with field data collection teams to solve infrastructure and pipeline bottlenecks while ensuring reliable production operations.

Responsibilities

• Design and build the ingest path from capture rigs to cloud storage resilient to field network failures.
• Build transcoding and normalization pipelines to convert heterogeneous rig output into consistent media.
• Time-synchronize multi-camera video with telemetry and sensor streams into labeled episodes.
• Build automated quality checks to flag unusable sessions like blur, exposure failure, or dropped frames.
• Design dataset versioning, lineage, and indexing so shipped episodes trace back to raw captures.
• Own pipeline costs including storage tiering, lifecycle, egress, and GPU utilization.
• Improve pipeline observability through monitoring, logging, tracing, and metric tracking.

Requirements

• 5+ years building and operating production video or media processing systems at scale.
• Hands-on production experience with FFmpeg, GStreamer, or equivalent media frameworks.
• Fluency in codecs and containers such as H.264/HEVC/AV1 and MP4/fMP4.
• Strong understanding of distributed systems, cloud infrastructure, object storage, and observability.
• Strong programming and problem-solving skills.
• Experience with durable workflow orchestration tools like Temporal, Conductor, Step Functions, or Airflow.
• Experience designing cost-efficient architectures for large media or data workloads.

Good to have

• Experience with multi-camera, multi-view, or time-synchronized capture systems and clock-drift problems.
• Familiarity with robotics or sensor data formats like MCAP, rosbag, ROS/ROS2, or Parquet.
• Familiarity with machine learning on video or images including classification, detection, and segmentation.
• Experience with perceptual quality measurement such as VMAF and encoding-ladder design.
• Experience with streaming and packaging formats including HLS, DASH, CMAF, and DRM integration.

Who should apply

This role suits experienced software engineers with a strong background in media engineering, video processing, and distributed systems at scale. Candidates should have at least five years of professional experience building, scaling, and operating production video pipelines and cloud infrastructure. If you enjoy solving complex challenges involving video codecs, time-synchronization, workflow orchestration, and cost-efficient cloud architectures, and want to contribute to the physical AI and robotics revolution, you will thrive in this position. Freshers are not eligible as 5+ years of experience is required.

Frequently asked questions