Principal Data Architect, AI Operations

About the Role

As the Principal Data Architect for AI Operations at Agiloft, you will be instrumental in shaping the future of our data landscape, making it AI-first. You will own the entire data architecture for our Data Warehouse Foundation, ensuring it's optimized for AI consumption across various applications like GPT assistants, AI agents, predictive models, and operational intelligence, in addition to traditional BI and reporting.

Responsibilities

• Own the end-to-end data architecture for the Data Warehouse Foundation, designing for AI-first consumption across GPT assistants, AI agents, predictive models, and operational intelligence — in addition to BI and reporting.
• Lead data modeling across all 11 departments, designing canonical enterprise data models that serve cross-functional AI and analytics use cases without duplication or fragmentation.
• Design and implement the contextual intelligence layer — including RAG architecture, vector store strategy, knowledge base ingestion pipelines, and document and unstructured data processing — that powers Agiloft's enterprise knowledge system.
• Build and maintain the agentic data integration layer: real-time and near-real-time data access patterns, agent memory and state persistence design, orchestration data requirements, and agent output integration back into the warehouse.
• Own the AI/ML feature layer — feature engineering strategy and standards, training data pipeline design, feature store architecture, and model output integration — enabling predictive analytics across churn, pipeline health, and operational forecasting.
• Design and govern the operational data and GPT context layer, including structured context feed design for GPT assistants, data freshness and access SLAs for AI use cases, and cross-departmental data reuse standards.
• Lead the Data Warehouse Foundation build in partnership with the external consulting team — setting architecture standards, reviewing implementation against AI-first principles, and ensuring the five-wave build plan delivers a foundation that serves the full intelligence architecture.
• Design and manage data ingestion, ELT/ETL, and orchestration pipelines across all source systems, ensuring reliability, performance, and cost efficiency.
• Establish and enforce AI data engineering standards across the organization — prompt-adjacent data design, agent data access patterns, reusable pipeline components, and quality assurance processes.
• Own data access policy design and least-privilege access controls in partnership with Security, ensuring data made available to AI systems is governed, auditable, and compliant.
• Define data quality standards and monitoring processes for AI-consumed data, where quality failures have direct impact on model and agent performance.
• Partner with the Principal Data and Integrations Architect on infrastructure design, ensuring data modeling and AI consumption requirements are incorporated into pipeline and architecture decisions from the start — not retrofitted after build.
• Partner with the VP FP&A and Manager of BI & Data to ensure the semantic and metrics layers are technically sound and serve both AI use cases and reporting requirements.
• Manage the AI Ops data architecture roadmap, translating business and AI use case requirements from all 11 departments into sequenced, prioritized technical work.
• Maintain documentation and knowledge transfer standards for all data architecture, pipelines, and integration patterns — ensuring AI Ops-built infrastructure is reusable, auditable, and not dependent on any single individual.
• Collaborate with the AI Agent Engineer and GPT & AI Systems Lead to ensure data infrastructure supports agent orchestration, retrieval-augmented generation, and multi-step reasoning workflows.
• Define the roadmap for data science and AI data work in partnership with the VP of AI Operations. All roadmapping is managed within AI Operations.
• Evaluate and recommend data tooling, frameworks, and platform components in alignment with AI Ops' technology-agnostic, build-for-leverage approach.
• Other duties as assigned.