Senior Software Engineer: Data Platform

What we're looking for:

We need someone with 5+ years of experience building production software across backend, frontend, and data-heavy systems who has deep experience with video processing pipelines, signal processing at scale, and microservices architecture. You should be comfortable building your own solutions (rather than relying on off-the-shelf tools), leveraging AI to accelerate your work, and thriving in a fast-paced startup where you'll touch nearly every part of the stack. Bonus points if you have ML experience or have worked at an AI training data platform like Scale AI, Mercor, Surge AI, or Appen.

What you'll do:

- Build high-throughput video processing pipelines including transcoding, HLS packaging, depth-map rendering, frame-accurate synchronization, and GPU-accelerated media workflows at scale

- Take ownership of ingestion and processing pipelines for video, 3D assets, depth maps, telemetry, annotations, metadata, and model-generated signals

- Design and build event-driven workloads on Kafka and Kubernetes that move data through capture, processing, enrichment, QA, packaging, and delivery

- Build AI enrichment systems for scoring, metadata extraction, classification, quality review, redundancy detection, and dataset recommendations

- Build internal applications for capture teams, reviewers, operators, and data managers to track, inspect, improve, and assemble datasets

- Build customer-facing marketplace tools where AI labs can search, preview, evaluate, query, purchase, and receive datasets

- Design data models behind provenance, asset versioning, licensing state, QA state, delivery history, analytics, and customer entitlements

- Build monitoring and analytics to understand quality, throughput, bottlenecks, customer behavior, and dataset value

About Origin Lab:

Origin Lab is building the data infrastructure for world-model AI. We sit at the intersection of gaming and AI — acquiring, capturing, enriching, packaging, and delivering premium rights-cleared multimodal data from game worlds for AI training. We work with game studios (AA, indie, and AAA) to help them unlock a new revenue stream by providing their game data to AI companies for pre-training world models, robotics systems, and more. Backed by Lightspeed Ventures and SV Angel, we recently closed our $6M seed round and are scaling rapidly. Every engineer here shapes the product directly, reporting to the founding CTO with major ownership across the core platform.