I'm Victor. I'm a former strategy consultant turned data scientist with experience building machine learning platforms and projects across industries ranging from telecommunications, advertising and marketing, manufacturing, supply chain management, and logistics.
Visit my website to learn more or connect with me on LinkedIn.
My blog has been named one of the best mathematics blogs by Abakcus. Here are some of my recent writing:
- Logistics Center of Gravity Analysis in Python, published in Analytics Vidhya
- 10,000 Hours, Artificial Intelligence, and the Democratization of Expertise, published in Data Driven Investor
- AI Biases and Independent Variables, published in Data Driven Investor
Some projects I have worked on involved:
- Directing the $400MM USD network expansion of a North American telecommunications carrier by running a predictive big data simulation that leveraged a terabyte-scale database to forecast sales on the customer level and developed a detailed network plan with specifications down to last-mile connections to customer premises.
- Generating client sales lift of up to 13% by creating a large-scale media portfolio optimization program that was used across the Asia Pacific. The algorithm used was custom-built to accommodate large and complex problems such as multiple year cross-market and cross-portfolio optimization of media assets at a weekly level.
- Developing an automated flu-tracker that predicts flu incidence based on variables such as weather and search interest to help schedule marketing campaigns for cold and flu medicine. The flu-tracker was piloted across 5 markets in Australia and was subsequently rolled out to other Asia-Pacific countries.
- Successfully predicting the winning candidate in the 2016 Philippine national elections months before traditional analysis as part of a competitive analysis for a client political party. Self-developed econometric models used measured the effect of media and non-media factors on candidate popularity.
- Improving user interactions in a chat bot app for a multinational insurance company using insights from natural language processing, resulting in an active user count increase of 73%.
- Spearheading the data science and data engineering aspect of an analysis of social media user behavior over the Chinese New Year. This was published in the South China Morning Post.
- Improving the logistics efficiency of leading multinational sportswear brand by developing an SKU-level inventory prediction system for seasonal SKUs. The system uses K-Shape clustering to combine SKUs based on the shapes of their inventory curves before fitting regression models for each cluster.
- Developing the data science platform for a supply chain management company's costing center of excellence, providing merchandisers with insights for vendor negotiation and strategic sourcing.
- Applying operations research techniques to consolidate and optimize the brand-level supply chains of a leading Asian food and beverage conglomerate.