Director, Impact Engineering

ServiceNow ServiceNow · Enterprise · Dublin, Ireland · Engineering, Infrastructure and Operations

Director of Impact Engineering at ServiceNow, leading a team of ~30 engineers focused on AI-assisted guidance, recommendations, and proactive insights for customer success. The role involves engineering leadership, architectural direction for AI experiences combining platform capabilities with retrieval and multi-step user experiences, and defining quality/evaluation criteria for AI applications. Requires experience shipping production AI/data-intensive applications and managing managers.

What you'd actually do

  1. Engineering leadership. Building, developing, and retaining the team. The team has engineers at varying stages of AI engineering experience. The Director grows that capability across the team through technical leadership, architectural review, coaching, hiring, and the standards they set.
  2. Architecture and technical direction. Impact's AI experiences combine ServiceNow AI capabilities, retrieval and grounding over product and customer signals, state and context management for multi-step user experiences, and durable customer- and user-specific preferences. The work involves balancing routing, permission-aware retrieval, data boundaries, privacy, latency, cost, reliability, and output quality. The Director helps decide when to use platform capabilities, when additional product, application, or integration work is needed, and how to keep what ships secure, observable, supportable, and durable across releases.
  3. Quality and evaluation. Quality and evaluation are core parts of how Impact engineering groups build AI-assisted applications, including curated test sets, expected-output criteria, regression testing, safety controls, permission-aware retrieval, and failure-mode monitoring. For this group, the work involves defining quality criteria for its applications and designing measurement that catches regressions across model, prompt, retrieval, data, and product-experience changes.
  4. Cross-functional execution. The Director works closely with Product, Design, Architecture, Security, and AI platform teams to turn ambiguous product goals into shippable, measurable engineering plans.

Skills

Required

  • Engineering leadership of teams in the 25-50 range, including experience managing managers.
  • Experience leading teams that have shipped production AI, data-intensive, recommendation, search, or decision-support applications.
  • Strong judgment about when to use retrieval, workflow automation, recommendation logic, model calls, human review, or simpler deterministic product behavior.
  • A working understanding of how data quality, retrieval architecture, permissions, privacy, and evaluation design shape AI output quality.
  • Comfort with low-ceremony engineering operating models — small pods, short cycles, direct accountability.
  • 10+ years in enterprise software, with several years in engineering leadership.

Nice to have

  • LLM/retrieval-based product experience is strongly preferred
  • Experience building AI applications inside an existing platform's constraints rather than in a greenfield environment.
  • ServiceNow platform experience (Glide, Now Assist, AI Agent Studio) or comparable enterprise AI platform experience.
  • Experience leading a team with a mix of local and remote engineers in the EMEA region.
  • Experience with evaluation tooling for AI-assisted applications.
  • Experience building product experiences that use customer-specific context, preferences, or personalization while respecting privacy and governance requirements.

What the JD emphasized

  • production AI
  • LLM/retrieval-based product experience is strongly preferred
  • evaluation design
  • quality criteria
  • measurement that catches regressions

Other signals

  • AI-assisted guidance
  • recommendations
  • proactive insights
  • context-aware customer experiences
  • retrieval and grounding
  • state and context management for multi-step user experiences
  • quality and evaluation
  • measurement that catches regressions across model, prompt, retrieval, data, and product-experience changes