Lead Software Engineer - Agentic Systems

Mastercard Mastercard · Fintech · Pune, Mahārāshtra, India · Engineering

Lead Software Engineer to build and scale agentic systems, focusing on platform and product capabilities where AI agents orchestrate multi-step workflows, invoke tools/services, and operate under explicit guardrails. This role involves distributed systems engineering and agentic runtime design, including architecture, security, observability, and reliability.

What you'd actually do

  1. Own complex engineering problems spanning multiple services and teams; drive alignment and resolution across dependencies.
  2. Design and deliver secure, scalable, resilient backend services and orchestration layers that enable multi-step agent execution.
  3. Implement orchestration patterns for agents (e.g., supervisors/routers, stateful workflows, tool invocation frameworks) with deterministic controls and safe failure modes.
  4. Treat agent execution as first-class production behavior: ensure actions are attributable, authorized, auditable, and observable through logs/metrics/traces.
  5. Design guardrails and policy enforcement hooks that constrain tool usage and data access, and support reviewable “why/what happened” audit trails.

Skills

Required

  • Bachelor’s degree in Computer Science, Engineering, or equivalent practical experience.
  • Significant experience building and operating production-grade backend and distributed systems.
  • Strong proficiency in one or more backend languages (e.g., Java, Kotlin, Python) and modern service development practices.
  • Demonstrated ability to own a service/domain end-to-end: architecture, implementation, deployment, and operational health.
  • Deep understanding of secure service-to-service communication, authentication/authorization patterns, and secure coding practices.
  • Strong experience with observability and operations (monitoring, tracing, incident response, postmortems, reliability improvements).
  • Proven ability to lead through influence: aligning cross-functional partners, driving tradeoffs, and unblocking delivery across dependencies

Nice to have

  • Experience building systems that integrate LLMs, AI services, or agent frameworks; familiarity with tool invocation and workflow orchestration patterns.
  • Experience designing policy-driven platforms with auditability and compliance considerations.
  • Hands-on experience with cloud-native architectures and DevOps practices (infrastructure-as-code, automated delivery, secure runtime patterns).

What the JD emphasized

  • agent orchestration
  • tool invocation
  • multi-step agent execution
  • agent execution
  • tool usage
  • policy enforcement

Other signals

  • AI agents orchestrate multi-step workflows
  • invoke tools/services
  • operate under explicit guardrails
  • agent orchestration and tool invocation
  • agentic capabilities are secure, observable, reliable, and auditable
  • agent execution as first-class production behavior
  • policy enforcement hooks that constrain tool usage and data access