Pitney Bowes methodologies
summary
Pitney Bowes and Ironside describe a production pipeline: preprocess billions of spatial rows into a spatial data lake, join to 1,400 datasets via a persistent PB Key, and serve microservices (ten millisecond 10 ms geocode) for real-time use. They show concrete results: a 1,000-listing Airbnb model in DataRobot where prior-year bookings and text/neighborhood scores predict summer demand, a crime model combining weather and incidents that shifts patrol routes, and retail models where drive distance is the top purchase predictor.
AI-written summary · about
chapters (10)
- 0:00Introduction and speakers
- 4:07Pitney Bowes spatial solutions
- 7:24Big data GIS processing trends
- 13:54Data science and business outcomes
- 25:52Airbnb use case analysis
- 32:15Retail loyalty insights
- 35:25Crime prediction case study
- 38:13Data updates and PB key value
- 41:24Insurance risk modeling
- 42:00Q&A session
related talks
19mExplore geocoded datasets with PwC’s innovative data platform
22mTake your Business to a New Geography
31mUnpacking the geospatial engineering toolbox – an overview of data science techniques for spatial da
23mDelivering Retail Analytics & Data Projects With Data Created For Purpose
25mA Journey Through a Geospatial Data Pipeline: From Raw Coordinates to Actionable Insights
22mHow Deliveroo Use Spatial Data Science To Fuel Growth
22mPython Applications in Infrastructure Planning and Civil Engineering
26mWho Said Wrangling Geospatial Data at Scale was Easy?