Principal Software Engineer - AI-First Development

Job Description: Position Overview The primary responsibility of the Principal Software Engineer (AI-First Development) is to direct the day-to-day technical execution of a small AI-First engineering team, designing, orchestrating, and validating software applications built through AI-driven development workflows. This role operates within an AI-First Software Development Lifecycle (SDLC) in which AI agents serve as primary producers of code, configuration, and test artifacts, while the Principal Software Engineer provides architectural direction, context engineering, human-in-the-loop governance, technical mentorship, and final accountability for delivered software. The Principal Software Engineer is a seasoned engineer who has already integrated modern AI-assisted development tools into their daily workflow and who has experience guiding other engineers through architectural decisions, code reviews, and delivery commitments. All duties are to be performed in accordance with departmental and Las Vegas Sands Corp.’s policies, practices, and procedures. All Las Vegas Sands Corp. Team Members are expected to conduct and carry themselves in a professional manner at all times. Team Members are required to observe the company’s standards, work requirements and rules of conduct. Essential Duties & Responsibilities - Agent Workflow Design and Orchestration - Define, build, and maintain the AI agent workflows the team uses to produce application code, infrastructure configuration, test suites, and documentation, and guide other engineers in extending them. - Decompose application requirements into discrete, well-scoped tasks that AI agents can execute effectively within defined boundaries, and review task decomposition produced by team members. - Select and configure appropriate AI models, agent frameworks, and tooling for each workflow based on task complexity, risk level, and cost considerations, and set the defaults the team works from. - Construct and maintain shared context that provides agents with organizational knowledge, coding standards, architectural patterns, and domain information needed to produce correct and consistent outputs. - Own the team's agent toolchain, including reusable skills, automation hooks, MCP integrations, and project memory files that provide persistent context across agent sessions. - Apply scoped subagent patterns where appropriate, following the principle of least privilege for tool access, and coach engineers on when multi-agent architectures are warranted versus when simpler workflows suffice. - Systematically capture insights, patterns, and failure modes from each development cycle and encode them back into shared context, skills, and agent configurations so that subsequent work becomes more reliable. - Lead collaborative requirement refinement sessions to align the team on acceptance criteria and context packages before agent execution begins. - Verification and Quality Assurance - Apply and uphold a multi-layer verification approach to AI-generated outputs, validating functional correctness, security posture, performance characteristics, code quality, and regulatory compliance. - Set the human oversight expectations at governance checkpoints appropriate to the risk level of each workflow, including pre-execution review, in-flight observation, and post-execution audit, and verify the team is operating to them. - Serve as the final reviewer and approver of AI-generated code for non-trivial changes, ensuring it meets Sands coding standards, architectural guidelines, and security requirements before promotion to production. - Build and maintain automated verification pipelines that supplement human review, including test harnesses, static analysis gates, and runtime telemetry. - Identify and lead remediation of patterns of agent drift, hallucination, or quality degradation across repeated workflow executions. - Define the team's agent observability practices