Applied AI ML Lead

JPMorgan Chase JPMorgan Chase · Banking · Bengaluru, Karnataka, India · Asset & Wealth Management

Lead the design and delivery of agentic AI solutions to improve software quality, security, and resiliency in a regulated environment. Build and scale autonomous agents for code remediation, standards conformance, and operational insights, integrating with CI/CD pipelines and establishing robust evaluation programs.

What you'd actually do

  1. Own end-to-end technical direction for agentic capabilities: architecture, delivery plan, reliability, security, and adoption.
  2. Design and implement LLM-driven agents for code generation/refactoring, standards conformance, test creation, documentation updates, release readiness checks, and operational insights.
  3. Establish safe “auto-heal” patterns: diff/PR-based remediation, risk-tiered actions, human-in-the-loop approvals, and explainable decisions.
  4. Build orchestration and coordination for multi-agent workflows (e.g., LangGraph / AutoGen or similar): state management, tool-calling, structured outputs, and guardrails.
  5. Implement **outer-loop** pipeline agent stages as compensating controls: policy checks, risk scoring, exception routing, evidence collection, and release gating.

Skills

Required

  • software engineering concepts
  • technical leadership
  • LLM-enabled applications
  • agents/tool use
  • structured outputs/validation
  • grounding/RAG
  • observability
  • SDLC understanding
  • IDE/inner loop
  • PR workflows
  • CI/CD/outer loop
  • regulated environments
  • security-first engineering mindset
  • stakeholder communication

Nice to have

  • Python (FastAPI/Pydantic)
  • distributed systems/reliability
  • automated remediation
  • codemods/refactoring tools
  • policy/guardrail systems
  • explainability
  • observability discipline
  • logs/metrics/tracing
  • OpenTelemetry
  • common monitoring platforms

What the JD emphasized

  • agentic AI
  • autonomous agents
  • LLM-driven agents
  • multi-agent workflows
  • outer-loop pipeline agent stages
  • rigorous evaluation program
  • LLM-enabled applications
  • regulated environments

Other signals

  • agentic AI
  • autonomous agents
  • LLM-driven agents
  • multi-agent workflows
  • outer-loop pipeline agent stages
  • rigorous evaluation program
  • LLM-enabled applications