Generated insights from 20+ programmatic campaigns by analyzing log-level data using PySpark and Databricks.

Performed audience analysis, shopper segmentation, and sales trend identification, contributing to a 15% ROI uplift and spotlighting categories driving 40% of seasonal sales.

Engineered Python automation workflows using python-pptx to dynamically update weekly campaign reports, reducing manual effort by 93% and accelerating insights delivery for hospitality and education clients.

Collaborated with account managers to deliver persona and trend insights using shopper behavior data that shaped creative strategy and timing, driving a 20% lift in CTR and 12% VCR improvement across top platforms.

Managed 15+ Apache Airflow data pipelines, ingesting terabytes of data from 10+ vendors, including credit/debit card transactions, web traffic, location data, and financial metrics via API, S3, and SFTP.

Used Python, Apache Spark, and Databricks for transformations, achieving a 40% processing time reduction.

Created a retail-based consumer data dashboard with Apache Superset using SQL to develop virtual datasets for rolling calculations of key metrics, e.g., retail store foot traffic, sales, and app/website active monthly users at company and sector levels.

Architected a decision analytics pipeline using AWS stack (S3, Lambda, Glue, RDS) to ETL financial models to a bitemporal database for audits and downstream analytics.

Curated an email-alerting system providing a summary of the models for 1,000+ analysts using SMTP (summary included an automated quality check for formula-checking for key metrics and highlighting missing year-periods).

Successfully migrated the pipeline to an EKS-hosted framework as an individual contributor deciding all timelines, changes, and developing unit tests within 4 weeks.

Developed a Python application for end-to-end automation of monthly performance report data extraction from six portfolio companies of a private equity firm and populated the extracted data into Tableau for easy viewing and analysis, saving over 80 work hours per month.

Facilitated team efficiency through comprehensive documentation of pipeline tasks, average processing time, notebook details, and creating automated QC notebooks on Databricks.

Conducted 10 research visits to various Delhi slums to understand ground-level problems and current methods of dealing with them.

Researched product-based ideas (cost analysis, local market research, etc.) as an alternative solution to some problems faced by certain communities in and around Delhi.

Worked on cost analysis and procurement of material for a pilot project of an aloe vera soap recipe made of waste vegetable oil supplied by restaurants (visited restaurants near the college for that purpose).