Deep learning on transactional time-series data to optimize supply chains and maximize margins

Supply chains are 60--70% of the operating expenses for manufactures. Supply chain data feeds drives the product's cost, quality, time-to-market, competitiveness, and availability. Almost all supply chain data is structured as time-series transactional data. The external signals that impact supply chain for manufacturers is also often time-series data. This includes commodity prices in the open markets, weather data, GDP data, sales data and other economic data etc. In this talk we show how to frame such problems and structure the data so they are fit for deep-learning engines, and how to optimize the deep-learning models to deliver high accuracy decisioning.

Founder & CEO/CTO DeepVu