Data Analyst

Company Overview

YouTube has a vast creator base from individuals to the biggest media companies. They create all different types of content offerings including live streams, paid subscriptions and video on demand.

In this role, you will be working side by side engineering team who’s in charge of building search and recommendation feature, in order to evaluate and improve the quality of a newly built service. The role is fast-paced and will involve creating set of query terms, genre / category tagging of a content and also various tasks related to evaluating and ensuring the quality of content recommendation and search.

Job Summary

• Running regular manual evaluation of search results against set of test queries
• Regularly updating search queries from aggregated data
• Regularly scraping search results from set of test queries to update and re-run mTurk(mechanical Turk- Amazon Human Eval Platform) tasks and keeping track of measured search quality by release
• Running more complex analyses we develop against results

Responsibilities and Duties

• Creating golden user profiles to evaluate recommendation quality, and keeping track of recommendation quality for each release as well as running manual mTurk evaluation
  
  Helping build new mTurk templates for quality evaluation of our search and recommendation service

Qualifications and Skills

• BA/BS degree in media, operations or relevant field of study around internet service
• 3-5 years experience analyzing Data sets
• Loves watching TV shows, movies, YouTube content, and other long-form video content and good sense of breadth of popular content in the market
• Experience evaluating search terms, search quality or recommendation quality
• Good understanding of XML/HTML and Javascript
• Python/SQL knowledge is a definite plus.
• Have strong organizational skills to keep track of all golden queries and user profiles
• Experience creating human evaluation templates and using mTurk or equivalent framework
• Communicate effectively to a variety of audiences including operations, engineering, product management, and user experience teams
• Quick learner and self-motivated