Principal Infrastructure Engineer / Kubernetes, Terraform, Python / On-site / Up to $400k at TrustIn
This company is building a human foundation model — real-time emotional intelligence for AI. It's a fundamental and largely unsolved problem, and the infrastructure challenge that comes with it is just as novel: low-latency multimodal inference, real-time WebRTC video, and large-scale data pipelines, all running reliably at production scale. You'll own the serving stack, the GPU clusters, and the deployment systems that make it all work.
WHAT YOU'LL DO:
• Design and optimise the inference infrastructure for latency, throughput, and cost — for workloads that span video, audio, and language in real time.
• Build and manage systems for real-time, long-lived WebRTC connections — ensuring smooth, low-latency delivery at the quality a conversational AI product demands.
• Orchestrate robust pipelines for offline processing, evaluation, and training using frameworks like Dagster or Ray — the data infrastructure that makes model improvement possible.
• Configure, maintain, and optimise GPU clusters using Kubernetes and Terraform. Develop CI/CD, evaluation, and versioning systems for safe, zero-downtime model deployments.
• Collaborate closely with researchers and product engineers to build the foundational infrastructure layer — shaping technical direction in a flat structure where your decisions stick.
WHAT WE'RE LOOKING FOR:
✓2+ years building production-level ML infrastructure — you've taken systems from zero to production and kept them running
✓Experience with LLM inference systems or multimodal systems (video, audio, multimedia) — ideally both
✓Strong hands-on experience with Kubernetes, Terraform, and major cloud platforms
✓Proficient in Python and either Rust or Go — you write production code, not just config
✓Extensive experience building data pipelines as distributed systems — not just scripting ETL jobs
✓Based in Seattle— this role is fully in-office, 5 days a week