SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Twelve Labs is building AI infrastructure that enables machines to understand video at scale. Video comprises 90% of global data, yet most remains inaccessible to machine understanding. The company has raised over $300M from top-tier investors including NEA, Radical Ventures, Amazon, NVIDIA, and others, with advisors including Fei-Fei Li and Silvio Savarese. Twelve Labs operates globally from San Francisco, Seoul, New York, and London, with core R&D in Seoul.
As a Token Engineer, you will design and build systems to manage AI token usage, cost, and optimization across the organization. This is not about restricting AI use or post-hoc cost auditing—it's about enabling researchers and engineers to leverage AI actively while maintaining consistent control over cost, performance, security, and reliability.
You will own the full lifecycle from initial design through production operations. Key responsibilities include: designing and implementing instrumentation pipelines to collect token usage, costs, latency, errors, and metadata from multiple AI tools and model APIs into a unified data model; building real-time dashboards, automated alerts, and anomaly detection systems; designing and operating a central AI gateway with authentication, authorization, usage limits, budget controls, rate limiting, model routing, caching, and audit logging; defining rational baselines and policies that prevent excessive costs and operational risk without hindering productivity; designing efficiency metrics beyond raw token counts (cost per successful task, latency, failure rates); building secure, scalable services on AWS and Kubernetes with Infrastructure as Code, secrets management, RBAC, and retention policies; defining SLOs, runbooks, and incident response procedures; providing self-service tools and documentation for teams to understand their usage and optimization opportunities; evaluating new models and providers; and translating cross-functional requirements into technical designs while managing tradeoffs.
You will collaborate with Infrastructure, Security, Finance, IT, Research, Product, and Engineering teams to build a safe, transparent, and automated AI usage environment.
