Ajit Kumar

• Architected end-to-end cloud data pipelines on GCP, ingesting large-scale data from diverse source systems into enterprise-grade data lake and warehouse layers.
• Deployed Kafka Connectors for real-time Change Data Capture (CDC), enabling low latency, event-driven data streaming across distributed systems, aligned with real-time and near real-time analytical requirements.
• Engineered real-time stream processing infrastructure using Cloud Dataflow to ingest and normalize high-concurrency event streams into BigQuery staging layers, with zero data loss.
• Designed and optimized complex BigQuery Stored Procedures for downstream ELT transformations, moving data from staging to conformed enterprise datasets using dimensional modeling standards applicable to MPP systems, like Snowflake.
• Built dependency-aware DAGs in Cloud Composer (Airflow) to orchestrate Kafka streams, Dataflow jobs, and BigQuery transformations across multi-functional pipelines.
• Collaborated with cross-functional analytics, data science, and BI teams to design resilient data models, improving data quality and downstream accessibility to drive critical decisions across Retail, Operations, and Finance verticals.

• Designed and deployed modular dbt models for complex transformation logic, reducing analytical code redundancy by 35% and aligning data assets directly with corporate KPIs.
• Implemented rigorous data quality frameworks and automated unit testing, boosting data processing reliability and reducing downstream reporting errors by 25%.
• Leveraged Terraform for automated schema enforcement and table creation via Infrastructure-as-Code (IaC), eliminating manual DDL overhead and accelerating environment provisioning by 40%.
• Streamlined CI/CD workflows using Buildkite for automated dbt deployments, reducing production release cycles from days to minutes and achieving 99.9% build reliability.
• Built and optimized high-throughput streaming and batch pipelines within the GCP ecosystem (BigQuery, Pub/Sub, GCS), scaling data ingestion capacity by 2x to handle heavy data spikes.
• Partnered with product and global analytics teams to translate complex business logic into scalable dimensional models, accelerating self-service BI dashboard load times by 30%.

• Architected high-volume ETL pipelines using AWS, Kafka, Python, Pandas, and Snowflake, cutting processing time by 30%.

• Automated data ingestion from Outlook and SharePoint APIs with Python, saving 40% manual effort.

• Designed audit logging, monitoring, and alerting frameworks for proactive data quality management.

• Automated Power BI dataset refreshes via Airflow, reducing manual overhead by ~20 hours/month.

• Built CloudFormation templates for data infrastructure provisioning, enabling scalable and cost-efficient pipelines.

• Standardized business logic and transformations via dbt, improving maintainability and consistency across models.

• Built and maintained scalable ETL pipelines using SQL, Spark, and Databricks, integrating data from multiple enterprise systems into analytics-ready datasets.

• Designed dimensional data models (star/snowflake schemas) to support enterprise BI reporting and analytics.

• Optimized Teradata queries, reducing execution times by up to 25% for high-volume analytical workloads.

• Enhanced pipelines to support data governance, lineage, and auditability for compliance-driven projects.

• Collaborated with cross-functional stakeholders in the Modern Analytics Insights Team to deliver new features and improvements