Rankush

Senior Data Engineer with 5+ years building enterprise ETL/ELT pipelines and cloud data platforms, currently leading data engineering across 3 concurrent enterprise projects — including a Microsoft Fabric lakehouse for Coca-Cola's dispenser fleet and a churn-prediction platform processing 900+ tenant accounts. Led a pipeline re-architecture that cut daily bug volume from 15-20 to 4-5 over 6 sprints and identified Azure infrastructure inefficiencies that saved $850/month in cloud spend. Earlier at DXC Technology, designed and built an automated daily reconciliation system validating 15,000-50,000 insurance policy records against a legacy COBOL mainframe migration, catching 300+ discrepancies that manual review missed, and cut a nightly SQL batch job from 8-9 hours to under 2 hours.

Senior Data Engineer at AppInventiv (2025-08 – Present)

• Team Leadership: Lead data engineering across 3 concurrent enterprise projects, managing 2 junior data engineers, 4 data analysts, and 1 AI/ML engineer, while supporting backend teams on cross-functional requests.
• Coca-Cola Fabric Lakehouse: Architecting a Microsoft Fabric medallion pipeline (Bronze/Silver/Gold) processing dispenser telemetry on a daily T1 batch cycle — currently live across 150 of a planned 1,000+ dispensers, ~10-20GB processed to date.
• Reliability & Cost Engineering: Redesigned a client data pipeline using Celery-based task queuing and chunked (vs. bulk) data movement with granularity checks, reducing daily bug volume from 15-20 to 4-5 over 6 sprints; separately restructured Azure pod/replica configuration across non-production environments, cutting cloud infrastructure spend by $850/month.
• Automation: Replaced manual client onboarding with an automated onboarding and migration pipeline including alerting, eliminating a previously manual, error-prone process.
• RetainIQ — Churn Prediction & Win-Back Platform: Designed and built a multi-tenant churn prediction system (30/60/90-day risk scoring) across 900+ tenant accounts, paired with an automated campaign and win-back verification pipeline; in one tracked month the model flagged ~799 at-risk users representing ~$27K in exposure, and win-back campaigns recovered 234 users (~$5,680 saved).
• Automated Reconciliation System (self-initiated): Identified that manual daily validation of insurance data (20-30 policies/day, 100+ tables) couldn't scale against 15,000-50,000 policies processed daily during a legacy COBOL-to-cloud migration; designed and built an automated daily reconciliation pipeline comparing both systems end-to-end, catching 300+ data discrepancies that manual review missed, still in active use (v15) after handoff.
• Performance Optimization: Rewrote and re-indexed legacy SQL across 250+ queries and restructured a nightly batch job, cutting runtime from 8-9 hours to under 2.
• Technical Enablement: Delivered a 15-day Python and machine learning training program, upskilling 500+ engineers and reducing external vendor dependency by 60%.

Data Engineer & Python Developer at DXC Technology (2021-01 – 2025-08)

• Automated Reconciliation System (self-initiated): Identified that manual daily validation of insurance data (20-30 policies/day, 100+ tables) couldn't scale against 15,000-50,000 policies processed daily during a legacy COBOL-to-cloud migration; designed and built an automated daily reconciliation pipeline comparing both systems end-to-end, catching 300+ data discrepancies that manual review missed, still in active use (v15) after handoff.
• Performance Optimization: Rewrote and re-indexed legacy SQL across 250+ queries and restructured a nightly batch job, cutting runtime from 8-9 hours to under 2.
• Technical Enablement: Delivered a 15-day Python and machine learning training program, upskilling 500+ engineers and reducing external vendor dependency by 60%.

Bachelor of Technology in Computer Science – Amity University, MP (2017-06 – 2021-06)