Metis Data Science Speaker Series

Transcript

1. Ike Okonkwo - @ikeondata Senior Data Scientist @AdRoll 8.25.16 Metis Data Science Speaker Series
2. About Me • Data Scientist - Programmatic Optimization • AdRoll - Global Leader in Retargeting and Performance marketing • Data Science mentor and writer • Background • Physics / Electrical Engineering • Industrial & Systems Engineering
3. AdRoll • AdRoll maximizes what lies beneath the surface of your customer data, attracting new prospects, converting them into customers, and growing their value • 15 Engineering Teams • 4 Data Science Teams • OptiLab (skunkworks) - optimizing AdRoll’s Real Time Bidding Engine
4. Data Science @ AdRoll ETL Layer : Presto, R, Python Data Layer : s3, Hadoop, Postgres Presentation Layer : Tableau ML Layer : Python, R, bash, SQL SAAS Layer : API R&D : trailDB, Python, R, D, VW, C, bash, Java, erlang, UDFs, Convex Optimization, MAB, GA, Real Time Bidding, Feature Engineering, Large Scale ML models, Decision Bayesian Networks, ANN, etc • 4 DS Teams : DSE, OptiLab (skunkworks) , Supply, Product
6. Interview Tips • Become comfortable at data take home challenges • Understand the basics : ML, Prob / Stats, SQL, Algorithms, Product, Domain knowledge • Network : meetups, linkedin, etc • Build a portfolio of interesting data science projects • Understand how most of the major ML algorithms work under the hood and how to tune and scale them out • Become more visible : blog, contribute to open source ML, etc