Company: Disney Interactive Media Group
Location: Palo Alto, CA
Rimmi Amjad, email@example.com
Disney Interactive Media Group (DIMG) , the interactive entertainment affiliate of The Walt Disney Company, creates immersive, connected, interactive experiences across console, online, mobile and social network platforms to entertain and inform audiences around the globe. DIMG's mission is to deliver Disney content to fans, whenever and wherever they want it, through numerous interactive media platforms.
Disney Interactive Media Group operates five global product groups: Disney Interactive Studios produces console and handheld video games for the Nintendo Wii™, Nintendo DS™, PlayStation®Portable, Xbox 360® video game and entertainment systems, PLAYSTATION®3 computer entertainment system, and the personal computer; Disney Online produces Disney.com and a portfolio of leading lifestyle websites for families, including Family.com; Disney Online Studios develops online virtual worlds, providing connected game-play experiences for children around the globe; Disney Mobile brings the best of Disney content to the mobile web, smart phone applications, and mobile games; in addition, DIMG manages a Disney-branded mobile phone service in Japan in association with Softbank; and Playdom produces best-in-class content for the rapidly growing platform of casual games on emerging platforms and social networks such as Facebook and MySpace.
All five product groups work together to create a variety of connected, multi-platform entertainment experiences.
We're looking for an intellectually curious, creative and disciplined quantitative thinker to join DIMG's Business Intelligence team as a Scientist, Predictive Analytics. The person in this strategic high-impact role will work with engineers and business owners to prototype and deploy algorithms and quantitative models to improve monetization and operating efficiency across multiple areas of DIMG's games business (e.g., performance marketing, merchandising and pricing, advertising, and site personalization efforts).
- A well-rounded individual with an advanced degree (Ph.D. preferred) from a top institution in data mining, statistics, machine learning, operations research, or similar quantitative field
- 5+ years experience building and testing complex algorithms and models on Terabyte-sized databases to answer difficult research or business questions, ideally for web-based products - you've not only learned about the techniques in class, you've also successfully applied them to novel areas
- Experience focused on predictive modeling, machine learning, data mining, or related areas in AI or statistics ideally applying these tools across disparate domains (from econometrics to computational linguistics to customer retention modeling)
- Experience with online advertising optimization highly desirable
- Experience managing technical resources highly desirable Knowledge and skills
- Demonstrated ability to apply statistics and algorithmic expertise to solve real world problems
- Fluency in at least one mathematical / statistical programming language (e.g., R, SAS), 3GL (e.g., C++, Java), or scripting language (e.g., Python, VBA) - the more core software engineering skills, the better
- Proficiency in writing SQL queries and working with large-scale databases, and ideally experience with Big Data frameworks and tools such as Hadoop, Pig, Mahout, Lucene / Solr, etc.
- Mastery of core techniques within unsupervised learning (e.g., k-means clustering, principal component analysis, association rules), supervised learning (e.g., linear / logistic regression, neural networks, decision trees / CHAID, Support Vector Machines, Hidden Markov Models / Bayesian methods, etc), numerical approaches like Monte Carlo methods, economic tools such as discrete choice modeling, and systems-based approaches including BDI / agent-based modeling and/or social network analysis
- Excellent verbal and written communication skills, with a knack for explaining complex models to non-technical audiences Philosophy and approach
- Appreciation for working in an open, supportive, non-competitive team environment, in which ideas are freely exchanged and everyone is working toward the same goal of developing tools and products to drive value for our games
- Measures success objectively, e.g., the models with the most statistically significant lifts over a random control group win
- A preference for agile / rapid prototyping approaches to product development in which proof of concepts are quickly built and tested - it's OK to fail from time to time, as long as you fail fast!
- Willingness to attack the lowest hanging fruit and use "80 / 20" approaches when appropriate rather than worrying about optimizations with diminishing marginal value - you never let the perfect be the enemy of the good!
- Flexible approach to solving business problems, i.e. willing to borrow ideas from the literature, accept valid suggestions from colleagues and managers, or try completely new approaches when circumstances warrant.
- A keen eye and concern for data quality, embracing the principle of Garbage-In-Garbage-Out, and taking care to sanity check the input and output at every step of the process
- A focus on understanding the fundamental drivers of the observed phenomena whenever possible - no black boxes
- Ability to communicate complex results to less quantitatively sophisticated users
- Intellectually curious, analytically rigorous, hard-working and a good business intuition are musts!
- Apply advanced analytic techniques such as data mining and statistics to design, implement, validate, and refine mathematical models and algorithms to understand and optimize the value of user engagement with DIMG's games using player-level behavioral and attitudinal data
- Synthesize, document, and syndicate key results within the organization, with an eye to finding new applications for existing models and tools
- Continually identify high-impact opportunities for improvement (e.g., automation, optimization) in new and existing domains across DIMG's product line based on top business needs, surfacing the research questions that we should be asking next
- Work with engineering team to turn working prototypes into scalable production-ready code
- Design and analyze experiments that validate different optimization approaches
- Act as an expert and evangelist in areas of data mining, machine learning, statistics, and analytics
BUSINESS UNIT OVERVIEW
Playdom produces a diverse portfolio of casual games for the rapidly growing platform of social networks, including Facebook and MySpace. With over 47 million monthly users and #1 rated games, Playdom is one of the fastest-growing social game developers, offering high quality community entertainment experiences.
Rimmi Amjad, firstname.lastname@example.org