Nov 2023 – Present · 2 yrs 7 mos · Bengaluru, Karnataka, India · Hybrid

• Influenced a $1.6M quarterly increase in brand marketing investment by partnering with Account Managers and Strategy Leads to project sales lift, uncover untapped geo-level demand, and drive market-level budget optimization
• Designed and deployed automated dashboards to monitor eater penetration, cohort churn behaviour, and engagement trends, surfacing actionable insights that enabled recovery of $1M+ in under-utilised marketing budgets through precise reallocation.
• Built end-to-end analytics pipelines (SQL, uWorc, Python) that automated campaign performance tracking, reducing 150+ manual hours weekly and delivering insights across 25+ brands and 150+ store cohorts in APAC.
• Led incrementality and uplift measurement via A/B testing frameworks, contributing to refined audience targeting, offer strategies, and double-digit ROI improvement for Sponsored Listings and Custom Offer campaigns.
• Conducted deep-dive diagnostics on campaign underperformance by analyzing audience misalignment, geo-cannibalization effects, pacing inefficiencies, and promotional inconsistencies, directly informing optimization roadmaps.
• Contributed to team knowledge management through the creation of query templates, SOPs, onboarding guides, and deep-dive playbook —improving operational efficiency, onboarding speed, and team productivity.

Jul 2021 – Aug 2023 · 2 yrs 1 mo · Bengaluru, Karnataka, India

• 1. Built a dynamic time-series forecasting system that autonomously selected the optimal predictive model (Moving Average, SARIMAX, Triple Exponential Smoothing, and Facebook Prophet) based on historical performance and product behavior.
• ➤ Result: Increased forecast accuracy from 34–45% to 50–65%, reducing planning inefficiencies and driving measurable business value.
• 2. Developed a 5-year strategic demand forecast for seasonal and high-variance SKUs, supporting long-term product planning, pricing, and inventory optimization decisions
• ➤ Result: Forecasting strategy directly linked to a projected 50% reduction in excess inventory and 10–15% boost in operational efficiency.
• Engineered a machine learning pipeline for new product demand forecasting, applying feature selection and variable importance scoring using Random Forest Regressor to model cold-start SKUs (which comprised 60% of product portfolio).
• ➤ Result: Delivered an intelligent framework that improved forecast accuracy for new items, a critical need in high-churn product categories.

May 2019 – Jul 2019 · 2 mos · Bangalore Urban district, Karnataka, India

• Generated experimental datasets from over 300 concrete cubes and 10 beams using Acoustic Emission (AE) monitoring techniques, enabling the detection of micro-level material failures in real time.
• Applied Gaussian Mixture Models (GMM) and Support Vector Machines (SVM) to classify crack types into micro, macro, and transverse categories, based on AE signal characteristics like amplitude, duration, and frequency.
• Translated raw experimental data into actionable durability insights, facilitating better understanding of material behavior under stress conditions—ultimately informing design improvements and long-term performance projections for civil infrastructure projects.
• Collaborated in a cross-functional academic setting involving civil engineering, materials science, and applied machine learning, bridging the gap between physical experimentation and computational modeling.