Principal Software Engineer

** Overview*
* Principal AI Engineer - AI Engineering Execution Excellence (Microsoft Security)

Transform how AI engineers build, validate, and ship-using AI to increase throughput, quality, and customer impact.

Microsoft Security (MSEC) Getting Customers Ready for AI (GR4AI) team is seeking a Principal AI Engineer to lead execution excellence for the AI engineering team. This role will translate the team's vision and strategy into a disciplined operating system for building and shipping secure, enterprise-scale AI solutions with greater speed, quality, and predictability.

The central mandate is to increase engineering throughput through an AI-native, NPF-transformative development model. You will redesign how engineers discover requirements, design systems, write and review code, create evaluations, investigate defects, document decisions, and operate services-embedding AI assistance and agents throughout the lifecycle rather than layering isolated tools onto existing practices.

You will remain deeply hands-on while operating across organizational boundaries: establishing repeatable delivery mechanisms, removing systemic bottlenecks, creating reusable paved paths, and coaching engineers to work effectively with AI. In partnership with the team leader, who owns vision and strategy, you will turn priorities into executable plans and ensure the organization consistently converts ideas into trusted production outcomes.

** Responsibilities*
* Own Execution Excellence and Engineering Throughput

+ Translate the team leader's vision and strategy into clear engineering priorities, executable plans, milestones, decision points, and accountable delivery rhythms.

+ Instrument the end-to-end engineering system to identify constraints in planning, design, implementation, review, evaluation, deployment, and operations; use evidence to improve flow rather than optimizing isolated activities.

+ Own measurable improvements in cycle time, deployment frequency, work-in-progress, quality, reliability, and engineer time spent on differentiated work, while avoiding output metrics that reward activity over customer value.

Transform Engineering with AI-Native Practices

+ Redesign the software-development lifecycle around AI-assisted and agentic workflows for discovery, design, coding, testing, evaluation, security review, documentation, incident response, and service operations.

+ Build reusable agents, context systems, evaluation harnesses, and paved paths that allow engineers to move from intent to validated production changes with less friction and stronger safeguards.

+ Establish standards for human oversight, provenance, secure tool use, data boundaries, review depth, and verification so increased velocity does not compromise trust, maintainability, or engineering judgment.

Lead Hands-On Delivery and Continuous Improvement

+ Work alongside engineers on the highest-leverage problems-prototype AI-native workflows, review critical designs and changes, diagnose systemic failures, and remove blockers that impede delivery.

+ Create reusable frameworks, templates, automation, and engineering standards that reduce cognitive load and enable teams to deliver faster without compromising reliability, security, or responsible AI.

+ Run disciplined learning loops through delivery reviews, retrospectives, experiments, and decision records; scale proven practices and retire processes or tools that do not improve outcomes.

Convert Priorities into Customer Outcomes

+ Partner with product, customer, and field teams to break strategic priorities into thin, testable increments that produce early evidence and shorten time to customer value.

+ Coordinate dependencies and resolve execution tradeoffs across Security, Azure, AI, Data, Research, and Customer Experience while keeping teams aligned to the established vision and strategy.

+ Connect delivery measures to customer adoption, task success, security posture, quality, reliability, time-to-value, and responsible-AI performance.

Own Production Trust and Responsible AI

+ Establish rigorous evaluation and release criteria for model quality, groundedness, safety, fairness, privacy, security, and abuse resistance.

+ Build telemetry and feedback loops that connect system behavior to customer outcomes, enabling rapid detection, learning, and continuous improvement.

+ Lead technical response to high-severity issues and ensure learnings become systemic improvements in architecture, testing, governance, and operations.

What Success Looks Like

+ The team reliably converts vision and strategy into prioritized, executable work with clear ownership, rapid decisions, and predictable delivery.

+ AI-native engineering practices materially reduce cycle time and toil while increasing deployment frequency, evaluation coverage, quality, and production confidence.

+ Engineers spend more time on differentiated customer problems because repetitive work, context gathering, verification, documentation, and operational tasks are safely augmented or…