About the job
We're building purpose-built infrastructure for running AI agents. Unlike traditional web apps, agents run for long durations, collaborate asynchronously with humans and other agents, and need to survive failures mid-execution. LangSmith Deployments is the runtime that makes this work, with durable checkpointing, fault-tolerant orchestration, and horizontal scaling, deployed across cloud and self-hosted environments.
Responsibilities
- Design distributed queue and worker systems that handle concurrent agent execution, background tasks, and multi-agent coordination across horizontally scalable infrastructure
- Own core data infrastructure — state persistence, atomic job claiming, connection management, and schema evolution
- Collaborate on architectural decisions, ensuring solutions are scalable and robust.
- Ship resumable streaming infrastructure so clients can disconnect and reconnect mid-execution without losing state
- Instrument and monitor production systems — tracing, metrics, and alerting to keep the platform healthy
- Participate in on-call rotations and own incident response for the runtime
- Create and maintain technical documentation, including system design and operational runbooks.
- Contribute to and extend open-source LangGraph, which is used by thousands of developers to build agent applications
Qualifications
Minimum
- 4+ years of professional backend engineering experience
- Strong proficiency in Go and/or python
- Experience with distributed systems — conensus mechanisms, queueing, state machines, and/or workflow orchestration
- Experience with scaling and sharding databases in high throughput environments
- Familiarity with Kubernetes, infrastructure-as-code, and at least one major cloud platform
- Strong communication skills and ability to work cross-functionally on a small team
Preferred
- Strong familiarity with Kubernetes (K8s), Terraform (Tf), and other DevOps tooling is highly preferred