Software Engineer, Agentic Systems - Moveworks

ServiceNow•Mountain View, CA
•Onsite

About The Position

Build and own the runtime infrastructure that powers Moveworks’ AI agents, including systems that orchestrate, execute, and deliver agent responses to enterprise users in real time. This is a distributed systems engineering role, not an ML role. The agents can plan and execute multi-step workflows, call tools, wait for human input, and resume while maintaining correctness, observability, and low latency. The role involves developing an agent orchestration engine; distributed session management with leases, heartbeats, and crash recovery; event-driven messaging and real-time streaming; concurrent asynchronous task execution; observability infrastructure; and caching and state layers.

Requirements

  • Deep experience in at least three of the following areas: distributed systems; concurrent or asynchronous programming; event-driven architectures; database systems for infrastructure; observability; or gRPC/protobuf.
  • 0–2+ years building production backend or infrastructure systems.
  • Strong proficiency in Python or Go; ideally both.
  • Experience designing and operating systems that handle real traffic at scale.
  • Comfort working with ambiguity and novel problems without textbook solutions.

Responsibilities

  • Build and own an agent orchestration engine that manages long-running sessions across planning, execution, and user interactions involving multiple LLM calls and tool invocations.
  • Develop distributed session management using lease-based ownership, DynamoDB conditional writes, heartbeat protocols, and checkpoint-based crash recovery.
  • Build event-driven message pipelines using SQS FIFO queues, Kafka consumers, and real-time streaming with gRPC and Socket.IO.
  • Implement structured concurrency with Python asyncio TaskGroups for concurrent session tasks, using fail-fast semantics and graceful cancellation.
  • Develop observability infrastructure using OpenTelemetry, distributed trace context propagation across asynchronous boundaries, and custom span lifecycle management for long-running sessions.
  • Build caching and state layers using Redis and DynamoDB KV stores, including per-organization and per-bot scoping, batch-read optimization, and hot-reload configuration.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service