Company: Top Entertainment & Mass Media Client
Requisition Title: Data Scientist
Assignment Length: 2 months (Possible extension)
Location: Los Angeles, CA (on-site)
Rate: $73/HR (W2 ONLY, NO C2C)
Preferred Industry: Experience in the entertainment industry or with PVOD and PEST forecasting.
What are the main 3-5 responsibilities of this resource?
- Develop and implement a machine learning model to forecast consumer demand and spending for PVOD and PEST.
- Incorporate diverse features such as marketing, film quality, competition, and pricing into the model.
- Support early-stage and post-theatrical forecasting modes.
- Ensure model explainability for non-technical stakeholders.
- Provide guidance for integration of the model into weekly workflows.
- Utilize historical data on theatrical performance, film attributes, marketing, pricing, and competitive context.
- Deliver a predictive model, explainability package, accuracy assessment, and implementation recommendations within a four-week timeline.
What are the TOP Critical Skills or Must Haves skills a candidate should have on their resume in order to be considered?
- 1-2 years of Proven experience in machine learning model development and implementation.
- Strong understanding of forecasting methods and techniques.
- Proficiency in programming languages such as Python or R.
- Experience with data analysis and feature engineering.
- Ability to work with historical data and incorporate diverse features into models.
- Excellent communication skills to explain complex models to non-technical stakeholders.
- Strong problem-solving skills and attention to detail.
The successful candidate will be responsible for developing an advanced machine learning model to forecast domestic consumer demand and spending for Premium Video on Demand (PVOD) and Premium Electronic Sell-Through (PEST) at the individual film title level. This role will involve improving current box office-based methods by incorporating diverse features such as marketing, film quality, competition, and pricing.
Key Responsibilities:
- Develop and implement a machine learning model to forecast consumer demand and spending for PVOD and PEST.
- Incorporate diverse features such as marketing, film quality, competition, and pricing into the model.
- Support early-stage and post-theatrical forecasting modes.
- Ensure model explainability for non-technical stakeholders.
- Provide guidance for integration of the model into weekly workflows.
- Utilize historical data on theatrical performance, film attributes, marketing, pricing, and competitive context.
- Deliver a predictive model, explainability package, accuracy assessment, and implementation recommendations within a four-week timeline.
Required Qualifications:
- 1-2 years of Proven experience in machine learning model development and implementation.
- Strong understanding of forecasting methods and techniques.
- Proficiency in programming languages such as Python or R.
- Experience with data analysis and feature engineering.
- Ability to work with historical data and incorporate diverse features into models.
- Excellent communication skills to explain complex models to non-technical stakeholders.
- Strong problem-solving skills and attention to detail.
Preferred Qualifications:
- Experience in the entertainment industry or with PVOD and PEST forecasting.
- Knowledge of marketing, film quality, competition, and pricing factors.
- Familiarity with early-stage and post-theatrical forecasting modes.
Current forecasting relies heavily on theatrical box office performance, which is a strong signal but often fails to capture context—particularly when a film over-performs or under-performs relative to expectations. This project aims to build on existing internal work by incorporating richer feature sets, including marketing intensity, film quality and sentiment, competitive environment, franchise and talent effects, pricing, and consumer exposure, preferably using a machine learning approach.
Primary Goal: Deliver a model achieving high title-level accuracy, improving absolute percentage error, weighted accuracy, and stability across genres and revenue tiers.
Additional Goals:
- Support two forecasting modes: early-stage (limited attributes) and post-theatrical (rich attributes).
- Ensure the final model is explainable to non-technical stakeholders, with clear visual examples of how key features influence predictions.
- Provide integration guidance for weekly forecasting workflows.
We will supply historical datasets including theatrical performance metrics, film attributes, audience and quality indicators, marketing data, retail pricing, competitive landscape, and calendar context. Contractor may recommend additional data sources.
A. Predictive Model: A working model capable of forecasting weekly and cumulative consumer de