Location Atlanta, GA
Long Term Contract
| Additional Info regarding SO/R: | Data Engineer JDSr Data EngineerCustomer Data PlatformABOUT THE ROLEBuild the Pipelines of Customer TruthThe Data Engineer, CDP builds and operates the realtime pipelines behind TMobiles Customer Data Platform the system of record for 100M customer profiles, now moving to Google Cloud to deliver realtime customer context, nearrealtime signals, and sub50ms API access. Any agent, store associate, or digital channel should see a customers prior conversation and continue from there |
Must have
5+ years data engineering
2+ years real time pipelines
Google Cloud
Pub/Sub + Dataflow (Beam) 3+ years
BigQuery 3+ years
Cloud Spanner 1+ year
Streaming &ETL
Spark Streaming / Kafka 3+ years
CDC, SCD Type 2, idempotent ETL
Expert SQL + PySpark
HYBRID MIGRATION
Databricks + Delta Lake 3+ years
Snowflake + Azure (Data Factory, Event Hubs, Cosmos DB) 2+ years
PLATFORM OPS
Terraform, CI/CD, Composer/Airflow 3+ years
IAM, VPC-SC, observability
SERVING APIs
GraphQL / REST with p99 < 50ms
The Data Engineer, CDP builds and operates the real-time pipelines behind client's Customer Data Platform - the system of record for 100M+ customer profiles, now moving to Google Cloud to deliver real-time customer context, near-real-time signals, and sub-50ms API access. Any agent, store associate, or digital channel should see a customer's prior conversation and continue from there.
You will build how clickstream, care, store, and network data flows into Cloud Spanner (low-latency store), BigQuery (analytics), and the knowledge graph behind GraphQL and MCP servers - while migrating and reusing proven ETL from the Azure estate (Databricks, Snowflake, Data Factory, Event Hubs, Cosmos DB).
RESPONSIBILITIES
01. Real-Time Pipeline Engineering
Build and run NRT ingestion with Pub/Sub and Dataflow (Beam) into Spanner and BigQuery, with data contracts, schema evolution, exactly-once semantics, dead-letter handling, and replay
Implement Spanner loads and change streams for current-state customer profiles, tuning keys, batching, and hotspots to hold sub-50ms reads
02. Clickstream, CDC & Data Modeling
Implement reusable CDC, SCD Type 2, and late-arriving-data patterns in Beam, SQL, and PySpark; build clickstream ingestion at scale
Partner with NBA and T-Engage to cut signal latency to sub-five minutes; build BigQuery Spanner sync and reverse ETL
03. Azure-to-Google Cloud Platform Migration
Move data from Azure to Google Cloud Platform with STS and BigQuery Spanner sync, reusing Databricks, Data Factory, and Snowflake ETL; execute migration or retirement of Cosmos DB, Redis, Event Hubs, and Functions with dual-run, reconciliation, and rollback
04. Knowledge Graph, API & MCP Data Feeds
Build the pipelines that populate the customer knowledge graph on Spanner (entity and identity resolution, Customer 360) and feed the sub-50ms GraphQL API and the MCP servers on Cloud Run used by AI agents
05. Quality, Operations & Delivery
Implement data quality, lineage, and access controls (Dataplex, Unity Catalog, TISS-310) and IAM/VPC-SC/CMEK; ship on Cloud Run, Dataproc, Dataflow, and Composer with Terraform, CI/CD, monitoring, SLOs, and cost tuning
Own pipeline on-call and runbooks, join design reviews, mentor junior engineers, and partner with product, analytics, and AI/ML teams
Required Experience
Overall (5+ years): Bachelor's in Computer Science or related field (or equivalent); 5+ years in data engineering, 2+ building real-time pipelines in production.
Google Cloud data platform (3+ years): Production data pipelines on Google Cloud Platform, including IAM, VPC Service Controls, and cost awareness.
Streaming & NRT pipelines (3+ years): Pub/Sub, Dataflow (Beam), Kafka, Event Hubs, or Spark Structured Streaming with exactly-once semantics, watermarking, and clickstream ingestion at scale.
BigQuery (3+ years): Partitioning and clustering, reservations and cost control, Spanner federation, and BigQuery-to-Spanner (reverse ETL) data movement.
Cloud Spanner (1+ year): Schema and primary key design, indexes, hotspot avoidance, change streams, and bulk and streaming loads.
Data modeling & ETL (4+ years): Dimensional, Data Vault, and medallion models; idempotent ETL/ELT (CDC, SCD Type 2); expert SQL, Python, and advanced PySpark.
Databricks, Spark & Delta Lake (3+ years): Spark tuning, Delta Lake, Delta Live Tables, and Unity Catalog, to reuse and migrate existing ETL.
Azure & Snowflake (2+ years): Data Factory, Event Hubs, Cosmos DB, Redis, and Functions, plus Snowflake, to support a hybrid-cloud migration.
Serving & APIs (1+ year): Building data feeds for REST/gRPC or GraphQL services with p99 targets under 50ms; exposure to MCP, LLM/agent, or Vertex AI data access a plus.
Platform engineering (3+ years): Terraform/IaC, CI/CD, Cloud Run, Dataproc, and Composer or Airflow; observability and SLOs.
Testing & data quality (2+ years): Unit, integration, and data-quality testing for pipelines; reconciliation, lineage, and Dataplex or similar tools.
Collaboration (2+ years): Working with product, analytics, and AI/ML teams; clearly explaining design tradeoffs; mentoring junior engineers.
