Principal AI Platform Engineer

ServiceNow•Santa Clara, CA

1d•$221,200 - $387,100•Hybrid

About The Position

The Principal AI Platform Software Engineer acts as the Forward Deployed Engineer (FDE) lead architect directing a group of FDEs in the CRM & Industry Engineering organization to design, build, ship, and operate autonomous intelligence systems — agent runtimes, reasoning pipelines, orchestration platforms, and AI-native applications. Across the full development and deployment lifecycle, the Principal FDE translates customer and product problems into precise specifications, directs autonomous and semi-autonomous agents to generate AI Agents, engineers the context and guardrails that make that output reliable, and verifies that the resulting systems are correct, secure, reliable, and maintainable.

Requirements

A demonstrated track record of building, shipping, and operating production software, with hands-on experience delivering real systems using AI-native methods (agentic coding tools and AI-assisted workflows) and working with customers in deploying AI as a Forward Deployed Engineer (FDE).
Bachelor's degree in Computer Science, Software Engineering, or a related technical field, or equivalent practical experience. Advanced degrees or relevant certifications are a plus but are not a substitute for demonstrated ability to ship reliable software with AI-native practices.
Software engineering fundamentals. Strong command of data structures, algorithms, system design, testing, and modern software development practices.
Solid AI/ML fundamentals. A working command of the machine learning concepts that govern how these systems behave — model training and evaluation, embeddings, and the probabilistic, non-deterministic output and failure modes of modern LLMs and deep learning
Fluency with AI coding tools and agent patterns. Practical, current proficiency with agentic coding tools and AI-assisted development workflows, including an understanding of agent architectures and tool-use patterns.
Agentic AI and autonomous systems. Hands-on experience designing and operating autonomous agent systems — planning loops, dynamic tool invocation, memory management, execution policies, and multi-agent collaboration — that select and sequence actions reliably at runtime, not in prototypes.
Context engineering. Skill in designing systems that determine what information models receive, when it is retrieved, and how context evolves over time.

Responsibilities

Act as the Lead Architect for a group of FDEs. Establish FDE best practices and reference architecture across the group of 30+ FDEs in the CRM & Industry Engineering organization deploying Agentic AI solutions across Case Management and Conversational channels like Voice and Chat.
Translate customer use cases into precise specifications. Convert customer AI use cases into efficient Agentic AI solutions using the ServiceNow AI Platform and other AI tools.
Orchestrate AI agents to build and refactor software. Decompose work into agent-sized tasks and direct AI coding agents and tools to generate, modify, and refactor production code across multiple files and services, supervising several workstreams in parallel.
Engineer context for reliable agent output. Author and maintain the context that drives correct results — system prompts, agent configuration and instruction files, architectural decision records, glossaries, golden examples, and agent-readable documentation — and curate what information enters the model at the right level of detail.
Verify and review output against intent. Rigorously evaluate code — human- or agent-generated — for correctness, spec adherence, security, performance, and maintainability, confirming behavior matches intent and that edge cases, failure modes, and concurrency are genuinely handled rather than superficially passing tests.
Make architectural and design decisions under uncertainty. Determine what to build, in what sequence, and within what constraints before agents begin work, and decide when to delegate to an agent versus implement directly, ensuring solutions fit cleanly into the broader system.
Own quality, security, and reliability in production. Take end-to-end responsibility for the software you ship — integrating changes safely, monitoring production, defending against risks like prompt injection and secret leakage, and feeding production signals back into specs and evaluation sets.