Software Engineering Manager , AgentOps
HybridBengaluru, Karnataka, India
Job Summary
Lead and grow a team of AgentOps Forward Deployed Engineers by hiring, coaching, and setting clear goals to ship high-quality agentic AI systems. Build and ship production-ready platforms by designing systems, conducting code reviews, and taking use cases from prototype to deployment with reusable connectors. Operate end-to-end observability for crews, instrumenting traces and cost metrics while establishing guardrails, identity models, and risk controls for enterprise integration. Collaborate with product, research, and engineering leaders to translate business needs into acceptance criteria and influence senior leadership on operational risk. Stay hands-on with Python and cloud-native backend engineering to set standards for agent orchestration, LLMOps, and multi-agent frameworks.
Required Qualifications
- Has led, coached, or mentored engineers - as a manager, tech lead, or team lead - and is ready to own hiring, performance, and growth for a team.
- Builds LLM-powered systems, AI agents, or workflow automation in production, and stays hands-on with the code.
- Strong Python and backend/platform engineering on cloud-native, distributed systems.
- Have worked on agent orchestration or multi-agent frameworks (e.g. CrewAI, LangGraph, ADK) and LLM observability tooling.
- Expertise in one of the public cloud platforms (AWS/Azure/GCP)
- Understands the agent lifecycle end to end - orchestration, identity/auth, observability, guardrails, and operating AI reliably in production.
- Sets standards and designs with clear trade-off awareness across quality, latency, cost, resilience, and maintainability.
- Excellent communication and stakeholder-engagement skills, from hands-on engineers to senior leadership.
- Bachelor's or Master's in Computer Science, AI/ML, or a related field, or equivalent practical experience.
Desired Qualifications
- Experience productionizing Agentic Applications
- Versed with MCP, A2A ,context engineering, and HITL agent execution.
- AI observability and evaluation frameworks (Langfuse / LangSmith / Arize Phoenix).
- Fluent in AI-pair-programming (Cursor, Claude Code) as a delivery mode, and able to raise a team's floor with it.
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.