Senior Data Architect

About the Company

The Senior Data Architect is responsible for defining and delivering enterprise-grade data architectures and reference implementations across complex, regulated environments.

About the Role

You will design scalable, secure, and cost-efficient data platforms and products spanning batch and real-time processing, lakehouse/warehouse, governance, lineage, and master/reference data. You will also enable GenAI/ML outcomes on governed enterprise data (RAG, semantic search, AI-driven insights, and agentic automation), partnering with Engineering, Security, Infrastructure, BI, and Data Science teams to translate business needs into target-state architectures, standards, and execution roadmaps.

Responsibilities

• Architecture, Standards & Roadmaps
• Define target-state enterprise data architecture and multi-year modernization roadmaps (cloud/hybrid), including migration waves and operating model impacts.
• Produce high-quality architecture artifacts: logical/physical models, integration patterns, data flows, domain/data product designs, and non-functional requirements.
• Own the architecture blueprint end-to-end (ingestion storage transformation serving) for analytics and AI.
• Set engineering standards and define reusable reference architectures and frameworks, templates and accelerators:
• Medallion patterns (Bronze/Silver/Gold) or equivalent layered curation
• Data products, domain boundaries/ownership, and contract-first delivery
• Data contracts, schema evolution/versioning, and deprecation/backward compatibility
• Ingestion, Integration, CDC & Streaming
• Standardize ingestion/integration patterns across API, batch, file, and CDC (where applicable).
• Architect streaming/event-driven platforms at scale: Kafka/Event Hubs/Kinesis + Flink/Spark Structured Streaming/Kafka Streams (or equivalents).
• Define reliability patterns: idempotency, dedupe, watermarking/late data handling, replay/backfill, and operational runbooks.
• Modeling, Curation & Serving (BI + AI)
• Model and curate data for consumption: curated datasets, dimensional marts, and semantic alignment for BI; fit-for-purpose datasets for ML/GenAI.
• Define serving patterns: governed SQL, semantic layers, APIs, and activation/reverse ETL (as needed).
• Establish performance/cost standards: partitioning/clustering, compaction, file sizing, workload isolation, and unit-cost/FinOps guardrails.
• Standardize storage/table tech: Parquet/ORC and Delta/Iceberg/Hudi.
• GenAI, RAG & Agentic Solutioning
• Solution agentic workflows for enterprise automation using Google ADK and/or AutoGen:
• Tool/function calling, planning vs execution, memory patterns, and human-in-the-loop approvals
• Guardrails: prompt-injection defenses, least-privilege tool access, grounding/provenance patterns, safe fallbacks
• Define scalable patterns for RAG/semantic search:
• Document ingestion (normalize/classify/dedupe), chunking strategy, metadata enrichment/ACL tagging
• Embeddings pipelines, vector index/retrieval, and secure context delivery
• Leverage platform-native AI where appropriate:
• Databricks (e.g., Mosaic AI, model serving; agent accelerators such as Agent Bricks where applicable)
• Snowflake (Cortex, Snowpark integration, governance-aligned AI consumption)
• Gemini Enterprise / Google AI suites (where applicable to client standards)
• Knowledge graph / semantic modeling implementation (entity resolution, taxonomy/ontology alignment; hybrid graph + vector retrieval).

• MLOps / LLMOps / AgentOps Enablement
• Establish production practices: reproducible training/serving datasets, registry integration, CI/CD for data + ML/GenAI pipelines, promotion across environments.
• Define evaluation + release gating: groundedness/quality metrics, safety checks, regression tests, monitoring/drift signals, and cost/performance baselines.
• Support model hosting/inference patterns (batch + real-time) and operational monitoring.
• Governance, Security, Compliance & Run-Ready Ops
• Define/enforce standards for modeling, quality SLAs/SLOs (freshness/latency/completeness), metadata/catalog/lineage, and auditability.
• Implement compliance-by-design controls: RBAC/ABAC concepts, row/column security, masking/tokenization, encryption, retention, and private connectivity patterns.
• Extend governance to AI usage: approved datasets for AI, access-controlled retrieval, and auditable context usage (per policy).
• Improve run readiness: observability (freshness/latency/failures), alerting, incident response, DR/backup/restore, and RPO/RTO awareness/testing approach.
• Architect and guide implementation of governance platforms (catalog, lineage, stewardship workflows) and ensure alignment with regulations (GDPR, CCPA, HIPAA as applicable).
• Master & Reference Data (MDM)
• Architect MDM and reference data solutions including domain ownership, golden record strategies, survivorship rules, and integration patterns.
• Define how master/reference data is published/consumed across analytical and operational systems.
• Stakeholder Partnership & Technical Leadership
• Act as a primary technical advisor; lead architecture reviews, mentor senior engineers, and drive cross-team alignment and decision-making.
• Partner with Security/Platform/BI/Product to ensure coherent enterprise-wide solutions.