Mumbai, Maharashtra, India
Summary
As a Data Engineer at Reliance Jio, I orchestrate large-scale data ingestion and processing, optimizing pipelines and delivering actionable insights for network reliability and user engagement.
Highlights
Orchestrated ingestion of 2TB UBR data from Oracle DB to Hive using Sqoop, optimizing pipeline runtime by 93% (from 10 hours to 40 minutes) and enabling faster identification of high-congestion building IDs.
Processed 50GB+ per partition of customer-level network data using Apache Spark (Scala) with structured Hive partitioning, identifying that 12% of cells caused 80% of outages, significantly improving network reliability.
Correlated 50GB+ batch customer ID data with cell names using Spark and Hive to identify high unavailability regions, automating root-cause analysis pipelines and reducing Mean Time to Resolution (MTTR) by 72% (from 18 hours to 5 hours).
Enabled tower health and coverage analytics for 200K+ assets, delivering actionable insights to 500+ cross-functional stakeholders across network, planning, and operations teams using AWS and Snowflake.
Enhanced the Apex DECK dashboarding suite by integrating daily, weekly, and monthly insights from STB, OTT, and Jio Gaming platforms using Snowflake, PySpark, and ADF pipelines, resulting in a 30% uplift in active user engagement.
Processed over 10M daily STB logs using PySpark to compute engagement KPIs, improving data quality by 40% and enabling accurate customer behavior insights.
Developed customer segmentation models from log analytics, revealing a 20% higher adoption rate of OTT platforms among users watching over 5 hours of YouTube per week.
Automated KPI reporting pipelines with Airflow, saving 20+ analyst hours monthly and reducing heatmap navigation time by 30% through optimization.