Principal Software Engineer

ServiceNow•Santa Clara, CA

10d•$221,200 - $387,100•Hybrid

About The Position

As a Principal Engineer in HRSD, you will set the technical direction for model-driven capability across the HR domain. This includes agentic and conversational experiences that interpret employee, manager, and HR agent intent, reason over profile, case, catalog, policy, and knowledge context, and act on the user's behalf. Your unit of ownership is the architecture, standards, and evaluation infrastructure that allow multiple teams to ship these experiences safely, rather than a single feature set. This role is not an ML research position focused on training foundation models, nor is it traditional full-stack work with deterministic logic. The job is shaped by two key aspects: much of the load-bearing logic resides in natural language (instructions, prompts, context, tool descriptions, guardrails) and must be engineered with the same discipline as code; and because behavior is probabilistic, correctness is established through large-scale evaluation, not fixed assertions.

Requirements

10+ years of software engineering experience, with a proven record of technical direction adopted beyond your immediate team.
Direct, hands-on experience authoring agentic instructions and prompts, designing autonomous workflows, and building the evaluation that verifies them.
Production experience with LLM APIs, retrieval-grounded features, agent orchestration, tool and function calling, and structured output enforcement, including at least one model or orchestration migration carried through production.
Entitlement-aware data handling: per-user access enforcement in retrieval and tool layers, scope separation between roles, and handling of restricted records under audit.
Evaluation experience used by engineers other than yourself: dataset curation, judge calibration, CI gating, drift detection.
Strong command of system design, APIs, data modeling, and testing across front-end, server-side, and relational data work with a high-quality mindset.
Effective, accountable use of AI coding assistants and agents.
On-call and incident-command experience on customer-facing systems, including ownership of the systemic fixes that followed.
Experience in software stacks including Java, JavaScript/TypeScript, React and demonstrated ability to quickly learn new tools and frameworks.
Bachelor's degree in CS, software engineering, or a related technical field, or equivalent practical experience.

Nice To Haves

HR domain depth: case management, employee journey and lifecycle, or payroll/benefits/leave/absence sufficient to set requirements.
Integration with core HR, payroll, or benefits systems of record, including reconciliation and eventual-consistency patterns.
Contribution in a Forward deployment model to champion customer adoption.

Responsibilities

AI architecture for the domain: defining how agents are decomposed and composed, where reasoning occurs, how context is assembled and bounded, how tools are exposed, and how autonomy is delegated.
Own multi-release calls: model selection and migration, orchestration approach, build-versus-adopt decisions, and the cost/latency/quality tradeoffs associated with them.
Define the autonomy boundary: determine as domain policy which HR actions an agent may take, which require a human decision point, and which no agent should attempt. Ensure that actions changing pay, employment status, restricted records, or employee-relations matters structurally require a human in the path.
Develop the shared instruction and tool description surface: treat this as a versioned contract with upgrade-safe extension points, deprecation paths, and compatibility guarantees, as customers configure, extend, and override it on their own instances.
Own evaluation as infrastructure: manage golden datasets, multi-turn suites, judge calibration, CI gates, and drift detection to enable teams to change behavior safely. Extend coverage to HR-specific failure classes like access-boundary violations, cross-scope leakage, and jurisdictional/policy-variant correctness.
Ensure production quality and safety: implement observability for containment, hallucination rate, tool-selection error, unsafe action, and injection vectors from user-supplied content entering agent context. Design diagnosis that works without exposing sensitive conversation content.
Establish AI-assisted engineering standards: convert ambiguous problems into testable specifications, define accountable agent-assisted delivery for the domain (specification standards, review expectations, verification harnesses), and enforce these standards.
Perform hands-on work where critical: tackle difficult integrations, risky migrations, prototypes that resolve architectural debates, and incidents that others cannot unblock.
Partner across functions: collaborate effectively with product, engineering, and design to co-create scalable AI solutions and translate business needs into robust technical designs.