Lead AI Products Architect

We seek a senior hands-on AI technical leader embedded within the AI Products Management Department. This role partners closely with Product Managers to translate product intent into scalable, secure, and production-ready AI system designs. Collaborate with the AI Research Team and the Delivery Team, the Lead AI Products Architect owns the technical quality bar across the AI product lifecycle – covering architecture, model performance, data readiness, and operational excellence.

This role does not replace Engineering or Research. Instead, it acts as the technical backbone of AI Products, ensuring solutions are measurable, reliable, and commercially viable in production.

• Own end-to-end technical architecture for AI products (GenAI, RAG, Agents, and ML), including inference services, orchestration, data pipelines, APIs, and integrations.
• Define reference architectures and reusable components (RAG modules, evaluation frameworks, prompt/versioning, feature stores, vector DB patterns, guardrails).
• Lead design reviews with Engineering, Platform, and Security teams to ensure scalability, resilience, cost efficiency, and maintainability.

• Define product-specific success metrics and acceptance criteria (accuracy, groundedness, latency, cost, safety, robustness, drift).
• Design offline and online evaluation strategies, e.g. A/B testing.
• Own the AI evaluation playbook (hallucination detection, retrieval quality, safety, jailbreak resistance; ML metrics such as precision/recall and calibration).

• Own AI product data strategy: sources, contracts, quality checks, lineage, governance, privacy-by-design.
• Define data readiness gates across PoC, MVP, and Scale phases.
• Partner with Data Engineering on pipelines and continuous data quality monitoring.

• Specify MLOps/LLMOps requirements: model registry, CI/CD, prompt/config versioning, observability, monitoring, rollback strategies.
• Define runtime controls including guardrails, policy enforcement, red-teaming, rate limiting, and audit logging.

• Act as technical co-pilot to Product Managers: translate product requirements into system constraints and non-functional requirements.
• Identify risks early and propose pragmatic trade-offs.
• Mentor Product Managers on AI technical fundamentals