VP, AI Data Architect
synchronycareers
Job Description
- Define the target-state Data & AI Foundations architecture supporting agentic AI use cases, including RAG pipelines, enterprise knowledge graph or metadata layer, data products, and AI-ready datasets.
- Own the strategy and roadmap for making key enterprise data sources "AI-ready": curation, quality, metadata, access patterns, latency requirements, and retention.
- Partner with source system owners (core servicing, CRM, collections, risk, fraud, etc.) to define data contracts, SLAs, and integration patterns that support downstream RAG and analytics.
- Design and govern canonical data models and semantic layers used by RAG pipelines, memory stores, and analytics to ensure consistency across agents and domains.
- Lead the design of RAG data infrastructure on cloud (e.g., PostgreSQL, Redshift, vector stores, object storage) and ensure it aligns with performance, cost, and compliance constraints.
- Define and implement RAG evaluation strategies including retrieval quality metrics, ranking and re-ranking optimization, relevance scoring, and A/B testing frameworks for continuous improvement.
- Establish data preparation and curation pipelines for model fine-tuning, including dataset selection, labeling strategies, quality validation, versioning, and compliance with model risk policies.
- Design and optimize retrieval strategies for RAG systems: chunking approaches, embedding models, indexing techniques, ranking algorithms, re-ranking logic, and hybrid search patterns.
- Build and maintain robust data pipelines (batch and streaming) that ingest, transform, enrich, and deliver data into RAG systems, vector stores, feature stores, and agent contexts with appropriate SLAs.
- Collaborate with the Enterprise AI Platform team on how data services (RAG APIs, feature stores, metadata services) are exposed as platform primitives for agent builders.
- Define and enforce data governance policies for AI: data classification, lineage, access controls, PII handling, retention, and usage logging for AI workloads.
- Partner with AI Governance/Model Risk and InfoSec/AppSec to ensure data usage in prompts, context, and tools adheres to policies, including regulatory, privacy, and model risk requirements.
- Establish data quality and observability practices for AI data: data SLAs, freshness, completeness, drift detection, and business rule validation tied to AI outcomes.
- Drive adoption of metadata and catalog tools so platform and agent teams can discover, understand, and safely consume datasets and RAG endpoints.
- Define and oversee patterns for integrating external data (third-party, public, partner data) into AI workflows, including licensing checks, quality assessment, and monitoring.
- Perform other duties and/or special projects as assigned.