Senior CRM Data Integration Architect

Warehouse, event-stream and API integration for lifecycle marketing platforms.

What you will own :

- Current-state discovery and source mapping. Inventory warehouse (Snowflake / Databricks), Kafka topics, APIs, microservices and legacy sources. Produce the source-of-truth matrix before taxonomy design starts.

- Target identity and relationship model. Define IDs and relationships across person, account/contract, tenant/brand and hierarchical structures. Design structural safeguards against cross-brand or cross-tenant data leakage, including single- vs multi-workspace decisions.

- Event and attribute taxonomy. Recommend the target schema, naming rules and payload requirements from a messy or partially documented starting point, working with the CRM strategy consultant.

- Consent and subscription architecture. Keep transactional, product and marketing consent states independent; design suppression, quarantine and consent-migration logic that survives legal review.

- Integration architecture and ongoing sync. Specify how each data class reaches the MAP from warehouse, stream or API : batch vs streaming vs reverse ETL, latency, deduplication, retries, error handling and reconciliation. Produce an ordered build sequence for the engineering team.

- Migration, backfill and implementation QA. Design historical import and cutover sequencing with acceptance criteria. Review build outputs, sample payloads and end-to-end MAP behaviour against the approved design, and sign off.

- Decision and blocker management. Maintain the decision/blocker log with owner, required decision date and impact. Chase client system owners for schemas, sample events, API specs and consent data.

- Hands-on build when required. Deliver connectors, middleware jobs, warehouse models and MAP ingestion when the client lacks capacity or a Mavlers build workstream is scoped.

- Pre-sales and practice. Scope and estimate the data/integration portion of proposals; join discovery calls; contribute reference architectures, QA checklists and connector patterns back to the BU.

Technical depth expected :

- Databricks : Delta Lake, Spark SQL/PySpark, Workflows, Unity Catalog; modelling and pipeline design, cost and performance. Depth : Deep

- Snowflake : Modelling, SQL, Streams/Tasks, Snowpipe, secure data sharing, warehouse sizing and cost control. Depth : Deep

- Kafka / event streaming : Topics, partitions, consumers, schema registry, CDC (Debezium/Kafka Connect); has shipped a streaming integration into a customer-facing system. Depth : Deep

- REST APIs, webhooks, export APIs : Design, build, secure and debug production integrations; pagination, rate limits, idempotency, retries; MAP export/data-feed APIs (Braze Currents, CIO Data Pipelines, Iterable exports). Depth : Deep

- One MAP data layer (Braze, SFMC, Customer.io, Iterable or Klaviyo) : Ingestion APIs, identity/profile model, event schema, subscription/consent objects, limits and quotas. Depth : Deep

- Remaining MAPs in scope : Know each platform's data model, identity rules and ingestion options well enough to blueprint and review. Depth : Working

- Salesforce (Sales/Service Cloud, Data Cloud) : Object model; REST, Bulk and Streaming APIs; Data Cloud ingestion and activation patterns. Depth : Working

- AWS : Lambda, S3, Glue, Step Functions, EventBridge, IAM; enough to build and review serverless middleware. Depth : Working

- CDP / reverse ETL : Segment, mParticle, Hightouch, Census : model-to-MAP sync, incremental logic. Depth : Working

- Middleware / iPaaS : Workato, MuleSoft, Boomi or equivalent; at least one in production. Depth : Working

- Languages : Python and SQL required; JavaScript/TypeScript working; dbt useful. Depth : Deep

Must-have requirements :

- 7+ years in data engineering or integration architecture, including 2+ years on customer/marketing data systems

- Hands-on production experience on Databricks, Snowflake and Kafka

- Has personally run discovery and written an integration blueprint that a client engineering team then built against

- Shipped at least one end-to-end warehouse or stream to MAP integration (Braze, SFMC, Customer.io, Iterable or Klaviyo)

- Has modelled identity across multi-entity or multi-brand structures and can explain how they prevented leakage

- Has designed or audited a consent/subscription model across transactional and marketing states

- Presents architecture to client engineering leads and defends trade-offs; writes specs engineers build from without a meeting

Nice-to-have :

- Direct Customer.io integration experience (Track/Identify API, Data Pipelines, Objects model)

- Databricks Data Engineer Professional, SnowPro Advanced, or Confluent certification

- Salesforce Data Cloud or Marketing Cloud Engineer credential

- Agency, consulting or SI background with multi-client context switching

- Privacy engineering exposure : GDPR/CCPA consent propagation across systems.