Kiran Kumar

Result-driven Snowflake data architect and senior data engineer with 14 years of IT experience, including 7 years specializing in architecting and optimizing the Snowflake cloud data platform. Expertise in transitioning enterprise infrastructures to modern data lakehouses, designing metadata-driven ELT pipelines, and building high-performance dimensional models. Proven ability to lead multi-cloud data migrations, implement cost-governance strategies, and mentor engineering teams. Strategic knowledge includes advanced Snowflake features such as dynamic tables, Iceberg tables, Snowpark native apps, and secure data sharing.

• Unilever, Engineering Specialist – Backup & Infrastructure, 01/18, 03/19
• IBM, Technical Services Specialist – TSM Administrator, 05/15, 01/18
• Adecco, Storage Administrator (TSM), 01/14, 01/15
• HCL Technologies, Technical Specialist, 02/12, 12/13

Project: Citi Bank – Legacy Hive to Snowflake Petabyte-Scale Lakehouse Migration
Role: Lead Architect / Senior Data Engineer

• Challenge: Migrating massive, legacy Apache Hive data structures with highly complex processing logic to an agile, cost-effective cloud warehouse framework without disrupting ongoing operations. The objective required ensuring absolute data integrity, sub-second latency for critical ledgers, and zero downtime for active transactional reporting streams.
• Solution:
• Architected and executed a secure, staging-to-analytics lifecycle architecture on Snowflake, migrating over 500+ operational transactional tables and petabyte-scale historical datasets from Hive, SAP HANA, AWS S3, and Azure Blob Storage.
• Engineered a highly resilient Audit, Balance, and Control (ABC) reconciliation framework to automate end-to-end ledger validation, ensuring zero data loss and achieving SOX and PCI-DSS audit compliance across all financial data pipelines.
• Designed a reusable, metadata-driven ELT framework using Snowpark Python DataFrames and Snowflake SQL, slashing pipeline engineering overhead by 85% and enabling automated schema evolution for runtime DDL changes.
• Enforced a zero-trust financial data security model by implementing multi-tenant Role-Based Access Controls (RBAC), Row-Level Security (RLS), Column-Level Security (CLS), and Dynamic Data Masking to protect Personally Identifiable Information (PII) and sensitive financial metrics.
• Deployed Snowflake Dynamic Tables to convert high-overhead batch processing into highly optimized incremental streaming pipelines, drastically lowering data freshness lag from hours to minutes for near real-time financial reporting.
• Formulated enterprise-wide dbt core standards, writing complex custom dbt macros, snapshots for Slowly Changing Dimensions (SCD Type 2), and automated data quality unit tests to maintain strict data lineage.
• Established a data platform FinOps program, leveraging advanced warehouse sizing automation, clustering keys, and the Search Optimization Service (SOS) to eliminate micro-partition churning and reduce compute spend by 30%.
• Implemented GitOps and Infrastructure as Code (IaC) via Terraform, embedding automated CI/CD deployment pipelines to safely provision multi-region Snowflake environments.
• Pioneered exploratory evaluations of open-table formats using Apache Iceberg and Snowflake Native Apps to future-proof storage flexibility and decouple compute dependencies.
• Impact: Decreased high-priority report generation latency from hours to minutes, achieved total automated data reconciliation with zero manual intervention, and minimized multi-cloud ingestion costs while upholding Tier-1 banking security protocols.