About
Data Scientist with 3+ years delivering CDP, Machine Learning, and LLM-powered GenAI solutions across Southeast Asia. Experienced in customer analytics, activations, customer unification frameworks, predictive models and demand forecasting for automotive clients.
Skills & Expertise (25)
Work Experience
Data Scientist
ADA
Jul 2023 - Present
Developed an XGBoost-based lead propensity scoring model on 70K+ leads, achieving 72% recall and 61% precision, enabling prioritized follow-ups and improved lead-to-sale conversion efficiency. Developed a CatBoost-based vehicle sales demand forecasting model using 3+ years of DMS data, engineering lag features, pricing signals, regional indicators, and other demand drivers to forecast monthly dealership-level sales. Achieved a final SMAPE of 23%, providing 3-month-ahead demand visibility to support production planning and inventory optimization. Designed and implemented a Treasure Data CDP solution using Medallion Architecture (Bronze โ Silver โ Gold), integrating data from 5+ sources including PostgreSQL, DMS, Web, and SFTP systems to build a unified customer data layer for analytics and activations. Built a deterministic Customer ID unification engine, created master customer profiles with derived and behavioral attributes, and created activation-ready segments to support marketing strategies. Developed 10+ dashboards across Sales, CRM, Aftersales and Overall Customer Lifecycle (OCL), providing actionable insights that enhanced operational monitoring and strategic decision-making.
Data Scientist
Korean Automotive Client
Present - Present
Developed an XGBoost-backed customer purchase propensity model within Treasure Data CDP to predict purchase likelihood and built an OpenAI GPT-4o-powered customer insight generator on top of the model outputs, providing instant 5-6 line customer summaries that eliminated manual data lookups, improved dealer productivity by 25%, and enabled more targeted lead engagement. Developed an AI-powered Prediction Monitoring and Root Cause Analysis Agent using Claude 4.5 Sonnet, automating daily monitoring of purchase propensity predictions, identifying key feature drivers, and generating business-friendly insights, reducing manual analysis time from 1 man-day to under 2 minutes.
CDP Delivery Lead
Leading Indonesia-based Automotive Resale company
Present - Present
Acted as CDP delivery lead across enterprise CDP and analytics engagements, despite a junior title. Independently delivered end-to-end CDP implementation for an automotive client, ran discovery workshops, including ELT development, ID-based customer unification, Master customer profile and segment creation, and WhatsApp Business Platform integration for automated reminders, and delivered web analytics dashboards. Built automated ingestion and transformation workflows, including ingestion alerts, automated data cleaning pipelines and data exports to ensure data reliability and monitoring. Designed and developed WhatsApp automated reminder workflows via WhatsApp Business Platform to drive customer re-engagement and retention. Activated CDP-defined audience segments across Google Ads and Meta (Facebook), enabling precision targeting and improved campaign performance.
Education
Bachelor of Engineering in ECE - Visvesvaraya Technological University
2019 - 2023 ยท Afghanistan
Certifications
No certifications added yet
Interested in this developer?
Profile Score Breakdown
Profile Overview
Skills (25)
Click a skill to find developers with the same skill