Principal AI Architect

Key Responsibilities Technical Ownership & Delivery - Own the AI dimension of assigned programs end to end, from requirement interpretation through architecture, build, integration, and go-live. - Prioritize and sequence use-case roadmaps based on business value and delivery feasibility, and set realistic, achievable delivery plans. - Act as the primary technical point of contact in stakeholder and working sessions, building and maintaining technical credibility and trust. Solution Architecture - Architect the data foundation (ingestion, storage, quality) required to support prioritized AI use cases. - Select and design the AI/ML approach for each use case, for example text classification, machine translation, retrieval and similarity search, LLM-based assistants or chatbots, forecasting, and decision-support scoring, based on business value and technical feasibility. - Design integration points into core business systems so that AI outputs are production-ready and directly usable in existing workflows. Hands-On Engineering & Delivery - Build priority models and pipelines directly, and direct any shared or contracted engineering support assigned to a program. - Establish MLOps practices (versioning, CI/CD, monitoring) sized appropriately to each program's scale and delivery cadence. - Validate delivered models against agreed success metrics and produce the evidence required for stakeholder sign-off at each milestone. Governance, Ethics & Compliance - Ensure AI components align with responsible-AI principles and data governance and privacy requirements, with explainability and audit logging built in by design. - Identify and escalate technical risks, gaps, and scope trade-offs promptly to relevant stakeholders and leadership. Stakeholder Management - Translate business and policy needs into scoped, deliverable technical work. - Communicate scope, timeline, and resourcing trade-offs clearly, and recommend additional resourcing where a workstream's scale calls for it.

Own the AI dimension of assigned programs end to end, from requirement interpretation through architecture, build, integration, and go-live. Prioritize and sequence use-case roadmaps based on business value and delivery feasibility, and set realistic, achievable delivery plans. Act as the primary technical point of contact in stakeholder and working sessions, building and maintaining technical credibility and trust. Architect the data foundation (ingestion, storage, quality) required to support prioritized AI use cases. Select and design the AI/ML approach for each use case, for example text classification, machine translation, retrieval and similarity search, LLM-based assistants or chatbots, forecasting, and decision-support scoring, based on business value and technical feasibility. Design integration points into core business systems so that AI outputs are production-ready and directly usable in existing workflows. Build priority models and pipelines directly, and direct any shared or contracted engineering support assigned to a program. Establish MLOps practices (versioning, CI/CD, monitoring) sized appropriately to each program's scale and delivery cadence. Validate delivered models against agreed success metrics and produce the evidence required for stakeholder sign-off at each milestone. Ensure AI components align with responsible-AI principles and data governance and privacy requirements, with explainability and audit logging built in by design. Identify and escalate technical risks, gaps, and scope trade-offs promptly to relevant stakeholders and leadership. Translate business and policy needs into scoped, deliverable technical work. Communicate scope, timeline, and resourcing trade-offs clearly, and recommend additional resourcing where a workstream's scale calls for it.