Databricks Architect

About the role

Role Summary: The Databricks Architect is responsible for designing, implementing, and governing enterprise-scale data platforms on Databricks. The role focuses on data architecture, cloud integration, data engineering best practices, analytics enablement, and platform optimization across Azure, AWS, or GCP environments.

Responsibilities

• Design and implement scalable data lakehouse architectures using Databricks.
• Define enterprise data models, governance standards, and security controls.
• Lead migration of legacy data warehouses and ETL workloads to Databricks.
• Architect batch and real-time data processing solutions using Spark.
• Optimize performance, scalability, reliability, and cost efficiency.
• Collaborate with business stakeholders, data engineers, data scientists, and cloud architects.
• Establish CI/CD, monitoring, and operational best practices for data platforms.
• Provide technical leadership and mentor engineering teams.

Requirements

• 8+ years of data engineering and data platform experience.
• 3+ years of hands-on Databricks architecture experience.
• Strong expertise in Apache Spark, PySpark, SQL, and Delta Lake.
• Experience with Azure Databricks, AWS Databricks, or GCP Databricks.
• Strong expertise in streaming platforms: GCP Pub/Sub, Azure Event Hubs, AWS Kinesis, Apache Kafka
• Hands-on experience with producer/consumer architecture decisions and trade-offs.
• Knowledge of data lakehouse architecture and data governance.
• Experience with data integration tools and orchestration frameworks.
• Strong understanding of security, access management, and compliance requirements.

Qualifications

• Databricks Certified Data Engineer Professional or equivalent.
• Experience with Unity Catalog, Delta Live Tables, and Databricks Workflows.
• Experience with Power BI, Tableau, or other BI platforms.
• Knowledge of DataOps, DevOps, and MLOps practices.
• Bachelor's degree in Computer Science, Engineering, Information Systems, or related field. Master's degree preferred.