Product Manager, AI Factories Orchestration, Agent Harness
Core
Build the Agent Harness for AI Factories orchestration, enabling LLM-based agents to safely operate AI infrastructure by defining tools, guardrails, and evaluation frameworks.
Role type
Senior Product Manager (AI Infrastructure & Agent Orchestration)
Builds
A platform connecting LLM-based agents to the DriveNets orchestration platform, including APIs, workflows, and runtime environments.
Domain
AI Infrastructure, High-Performance Networking, Cloud-Native Systems
Deliverable
production ML models
Required skills
LLM-based agent architecture (tool calling, RAG, MCP), AI infrastructure (GPU clusters, training/inference workloads), cloud-native software (Kubernetes, microservices), API design (REST/gRPC, OpenAPI), product strategy for developer-facing software, data-driven decision making, stakeholder management with hyperscalers.
Preferred skills
LLM evaluation (offline eval suites, regression benchmarks), agent frameworks (LangGraph, OpenAI Agents SDK), multi-agent systems, SRE/AIOps background, infrastructure-as-code (Terraform, Ansible), observability (Prometheus, Grafana, OpenTelemetry), on-premise/air-gapped deployment experience.
Technologies
Kubernetes, Slurm, NVIDIA Base Command Manager, Run:ai, NetBox, ServiceNow, Prometheus, Grafana, OpenTelemetry, gNMI, gRPC, Python, LangGraph, OpenAI Agents SDK, Claude Agent SDK, Terraform, Ansible, Helm, GitOps, CI/CD, REST, gRPC, OpenAPI, MCP, A2A.
Responsibilities
Define the Agent Harness tool layer and guardrails (RBAC, policy enforcement, audit trails); own first-party operations agents (incident investigation, change planning); build the evaluation framework for agent quality; design northbound APIs and SDKs for programmatic consumption; drive integrations with AI infrastructure ecosystem; run continuous discovery with hyperscalers and enterprise customers; create external technical content.
Seniority
Senior, hands-on IC