Yashasri Kanchukatla

· Designed AI-ready ETL/ELT pipelines processing 10TB+ daily on AWS (Glue, EMR, Kinesis, S3, Lambda) — reducing latency 35% and supporting real-time analytics and ML feature pipelines. · Built automated data quality and observability framework — 40% fewer discrepancies at 99.9% SLA — with metadata-driven validation, lineage tracking, and governance controls. · Optimized PySpark/Spark SQL and dbt transformation layers — improving processing speed 45% and cutting infrastructure costs 25% ($40K+ annually). · Implemented CI/CD (GitHub Actions) and Terraform IaC — enabling repeatable automated deployments and saving 100+ hours annually.

Architected Azure data lakehouse (ADF, Databricks, Synapse, ADLS Gen2, Delta Lake) supporting AI/ML workloads — improving pipeline performance 45% at 99.9% availability and 30% lower cost. Built dbt Medallion architecture (Bronze/Silver/Gold) with metadata management and governance — reducing dashboard load times 50% and delivering ML-compatible analytics-ready datasets. Delivered CI/CD automation (Azure DevOps) and Terraform IaC — reducing deployment failures 60%, manual effort 55%, with data masking, encryption, and anonymization across regulated environments.

Built real-time and batch pipelines on AWS (S3, Glue, Kinesis, Redshift) and GCP (BigQuery, Dataflow) with CDC and API integrations — reducing latency 45%, increasing accessibility 60%, cutting storage costs 20%.