Principal Engineer - Agentic AI Framework and Harness

Bank of America Bank of America · Banking · Charlotte, NC

This Principal Engineer role focuses on defining and leading the engineering approach for reusable agent frameworks and harness capabilities within the Software Delivery Lifecycle (SDLC) at Bank of America. The role involves shaping how engineering teams design, compose, test, and operationalize agentic workflows, requiring deep technical leadership in areas like agent orchestration, memory patterns, tool calling, planning loops, and simulator harnesses. The goal is to improve correctness, composability, governance, testability, and developer productivity for agentic systems in production within a regulated environment.

What you'd actually do

  1. Develops the engineering approach for the entire program/portfolio solution and works with Architecture, to develop/analyze/deliver the implementation of technical enablers
  2. Leads the planning, definition, and design of the complex features which span multiple teams and explore solution alternatives
  3. Creates ideas on designing complex technology and solution development approaches
  4. Leads the technical oversight for teams in solution development including design reviews and code within own domain
  5. Defines the technology tool stack for the solution within ranged of internally approved and supported technologies

Skills

Required

  • 15+ years of engineering experience with deep technical leadership in enterprise software platforms, developer tooling, or AI-enabled engineering systems
  • Demonstrated ownership of architecture, standards, and engineering direction for shared platforms used across multiple delivery teams or lines of business
  • Experience engineering solutions in highly regulated environments with strong SDLC, risk, audit, and control requirements
  • Ability to influence senior technology leaders, platform teams, and risk partners through clear technical strategy, architecture guidance, and engineering standards
  • Deep expertise in designing reusable agent orchestration frameworks, including execution models, control flow patterns, state handling, and composition strategies for multi-step and tool-augmented workflows
  • Strong comma

What the JD emphasized

  • highly accomplished
  • deep technical leadership
  • strong hands-on engineering depth
  • platform design expertise
  • track record of building reusable framework capabilities
  • deep expertise in designing reusable agent orchestration frameworks
  • strong comma

Other signals

  • Agentic AI Framework
  • Agent Orchestration
  • Reusable Agent Frameworks
  • Developer Tooling for Agents