Senior Software Engineer, Agents Platform

Box Box · Enterprise · Redwood City, CA · Core Platform

Senior Software Engineer to build the core platform for an enterprise-grade Agents Platform, enabling secure, reliable, and scalable AI agent development and operation. The role involves owning systems for agent frameworks, multi-agent workflow orchestration, LLM integration, and enforcing security/governance, with a focus on low-latency, high-throughput execution and observability.

What you'd actually do

  1. Build core components of the Agents Platform that power agentic use cases like Deep Search and Deep Research.
  2. Design and implement agent and tool repositories, along with observability and CI/CD pipelines, to streamline agent development and deployment.
  3. Develop and evolve a multi-tenant control plane that enforces isolation, fair resource allocation, and per-tenant SLAs across all agent workloads.
  4. Collaborate with ML engineers to translate requirements into scalable capabilities on top of LangGraph.
  5. Contribute to technical discussions and provide guidance on cross-team projects within the AI Platform organization.

Skills

Required

  • 5+ years of industry experience in computer science, machine learning or related field
  • strong understanding of distributed systems, data structures & algorithms, platform architecture
  • familiarity with at least one object oriented language like C, C++, Java, Scala
  • BS degree in Computer Science or a related

Nice to have

  • Advanced degree in computer science or related field.
  • Familiarity with concepts related to Large Language Models, Retrieval Augmented Generation (RAG) , Semantic Search, Indexing, Ranking and Relevance
  • Familiarity with cloud based ML platforms such as Vertex AI, AWS Bedrock, AWS Sagemaker etc
  • Experience with LangChain/LangGraph or other agent definition languages
  • Experience with Kubernetes bases systems

What the JD emphasized

  • build the core platform that makes agent development secure, reliable, and scalable
  • own systems that define agent frameworks and tooling, orchestrate multi-agent workflows, integrate with multiple LLMs and enterprise systems, and enforce tenant isolation, data governance, and least-privilege access
  • deliver low-latency, high-throughput execution with robust observability, guardrails, and safe execution environments
  • build the core components of the Agents Platform
  • Develop and evolve a multi-tenant control plane that enforces isolation, fair resource allocation, and per-tenant SLAs across all agent workloads
  • built, deployed, and supported distributed systems at scale

Other signals

  • building the core platform for AI agents
  • enabling agent development, deployment, and operation
  • integrating with LLMs and enterprise systems
  • enforcing security, governance, and isolation for agents