Member of Technical Staff
Chakra Labs · New York
About this role
## About Chakra Labs Chakra Labs’ mission is to encode human taste into intelligence. We build high-fidelity environments, evals, and datasets for frontier AI research—working with top labs. Our focus spans post-training, agent environments, data quality, and research infrastructure. We build systems that improve models in measurable, useful, and hard-to-fake ways. ## What You’d Work On - **Agent orchestration at scale**: Hundreds of concurrent agent runs, each with its own stateful environment. **100M tokens/minute** across the fleet. You’ll own the **dispatch layer** (SQS, concurrency control, failure handling). - **Environment and task design**: Create environments that feel real and scenarios that push agents to their limits. Build new evaluations and design tasks that test what matters—not just what’s easiest to measure. - **Product around the platform**: Infrastructure nobody can use isn’t infrastructure. Build customer-facing surfaces—**dashboards for run inspection**, **tooling for experts**, and **APIs** that make the platform feel obvious. Ship end-to-end. - **New frontiers**: The agent evaluation space is moving fast. Stay on the edge by supporting new environment modalities and shipping integrations with external orchestration frameworks. ## About You - **Generalist range + infra depth**: Strong across the stack (backend services, data pipelines, enough frontend to ship a real interface) with real systems depth. Prefer owning whole problems over single layers. - **Container orchestration**: Comfortable running **Kubernetes (or similar)** in production—auto-scaling, pod lifecycle, persistent storage, networking. Can debug scheduling issues and reason about resource contention. - **Distributed systems**: Experience with message-driven architectures (e.g., **SQS, Kafka**). Keep jobs moving under load, retry without duplicating, and fail without losing work. - **LLM infrastructure**: Run LLM workloads at scale—token instrumentation, rate limit handling, prompt caching, multi-provider routing. Build the plumbing between models and external tools and keep it stable under load. - **Experience**: Ideally **3+ years** at this level, but less may be fine if you match the above. ## What Makes This Different - **AI agents + infrastructure**: You’ll monitor model behavior alongside pod health, and debug token throughput alongside network throughput. - **Customers are researchers**: Work directly with labs pushing the frontier of what agents can do. - **Ownership, not theater**: Own whole systems, not tickets. One week you ship a new environment type; the next you scale dispatch to handle **10x throughput**. Build things that didn’t exist a month ago. - **Team**: Ex-Stripe, Snap, AWS, Microsoft, Airtable—small team with years of shipping high-impact products. - **Cutting edge**: Work across the latest data, AI, and infrastructure technologies.
Listing freshness
CronJobs last confirmed this listing 3h ago. If its source stops confirming the opening for seven days, this page is removed from active inventory.