SESSION

Leveraging Spark to Democratize Data for Omni-Commerce

Slides PDF Video

Insnap, a hyper-personalized ML-based platform acquired by The Honest Company, has been used to build a real-time data platform based on Apache Spark, Cassandra and Redshift. Users’ behavioral and transactional data have been used to build data models and ML models, and to drive use cases for marketing, growth, finance and operations.

Learn how Honest Company has used Spark as a workhorse for 1) collecting, ETL and storing data from various sources including mysql, mongo, jde, Google analytics, Facebook, Localytics and REST API; 2) building data models and aggregating and generating reports of revenue, order fulfillment tracking, data pipeline monitoring and subscriptions; 3) Using ML to build model for user acquisitions, LTV and recommendations use cases. Spark replaced the monolithic codebase with flexible, scalable and robust pipelines. Databricks helped The Honest Company to focus on data instead of maintaining infrastructure. While Honest users got delightful recommendations to improve experience, data users at Honest understood users much better in terms of segmenting with behavioral information and advanced ML models, leading to increased revenue and retention.

Session hashtag: #SFexp16

Shafaq Abdullah, Data Infrastructure at Honest Company

About Shafaq

A tech-savvy enterpreneur who loves to solve hard business problem with state-of-the-art technology. Currently Heading up Data Infrastructure at Honest after acquisition of Insnap- a BigData Learning platform as Cofounder and CTO. Leading Metric driven scalable data transport (real-time), processing and storage platform and services at GREE. Played key role in exit of Zenprise (Citrix)- by leading enterprise Mobile App management. Holds about 10 patents in Bigdata, ML, Entperise Security. MS from TUT, Finland and B.Sc Computer Engg from UET, Lahore.