Results-driven Data Scientist with a proven track record at Impact Analytics, specializing in machine learning and statistical analysis. Successfully delivered innovative SaaS solutions for clients like Ralph Lauren and Pacsun, leveraging advanced skills in SQL and pattern recognition to drive impactful retail strategies. Adept at transforming complex data into actionable insights.
PRICESMART-INITIAL PRICING-RALPH LAUREN, 10/2022-02/2024
-Developed COVID-adjusted initial pricing models at the product level while taking into account the historical data, current trends, and consumer behavior
-Worked in a highly dynamic environment, collaborating with various onsite and client teams to overcome the challenges in delivering the product
-Increased efficiency of the data pipeline by implementing and automating processes like data ingestion, storing secondary tables, etc
-Created initial price modeling based on product attributes using various regression techniques
-Implemented XGBoost modeling for initial price estimation, achieving a weighted absolute percentage error (WAPE) reduction of approximately 6% compared to alternative modeling approaches
-Created a price elasticity model based on sales patterns, which helped the initial pricing model for maximizing full-price sell-through of products
-Pricing of a product based on a similarity mapping model using product attributes as text embedding, and image embeddings
-Developed Python-based automations to reduce manual workload, optimizing efficiency, and saving 40 hours of manual effort per month
INVENTORYSMART-PACSUN, 02/2024-Present
-Analyzed and validated daily data received in Snowflake across multiple source tables before ingestion.
-Developed and orchestrated Airflow DAGs to automate the data flow from Snowflake to Google BigQuery, using internally built SelfHelp UI for trigger management and scheduling.
-Led the historic and periodic ingestion strategy, writing scalable SQL queries to support both one-time backfills and incremental loads across all source tables.
-Managed complete data flow through to master tables, and implemented DAGs to ingest processed data into PostgreSQL for downstream application use.
-Designed and scheduled backsync pipelines to re-ingest curated data from PostgreSQL back into BigQuery for analytics and reporting layers.
-Built and scheduled a weekly simulation process, automating a stored procedure in BigQuery that generates forecasting insights, integrated into a client-facing frontend tool.
-Ensured data pipeline reliability, quality, and scheduling accuracy, playing a key role in supporting business-critical weekly inventory forecasts for the client.
Particity-POC, 07/2023, 09/2023
- Cleaned the data and created master tables ready for modelling.
- Prepared and Presented Personal reports, analysis and presentations to various stakeholders.
Arhaus-POC, 01/2024, 02/2024
- Cleaned the data and created master tables ready for modelling.
- Prepared and Presented Personal reports, analysis and presentations to various stakeholders.