G2 The Principal Applied Scientist at G2 leads the development of ML systems for marketplace optimization and pricing strategies. This role requires deep expertise in empirical methods, Python programming, and advertising technology architectures.
Responsibilities
As a Principal Applied Scientist, you will build marketplace systems directly contributing to G2’s success. Your responsibilities will center around designing and implementing machine learning systems feeding a deterministic marketplace-optimization layer. Auction dynamics, pacing, bid rates+win rates, reserve pricing, budget utilization, and price discovery are first-order primitives in the modeling domain here.
Design+Implement ML Systems to drive marketplace outcomes (40%)
Reason about the metrology processes collecting training data and measuring response variables
Develop fundamental theories about the flow of information/causality through the marketplace modeling domain
Use this fundamental theory to inform ideas for predictive ML features and also to characterize the rough dimensionality of the problem and the approximate shape/complexity of the decision boundary
Have a sense of taste and intuition in hand-engineering predictive ML features
Reason about marketplaces from the first principles of economics (30%)
Deeply understand the relationship between prices, supply, and demand, in imperfect/lumpy real markets
Reason about nonlinearly-coupled marketplace dynamics over shorter and medium timescales
Expect significant second-order effects or longer-term feedback loops which aren’t illuminated by the shorter-timescale training data
Connect marketplace system interventions to the frontend UX and overall satisfaction of market participants (30%)
Strike a balance between following the data wherever it leads and being guided by a genuine vision for a compelling frontend UX
Qualification
Extensive experience with SnowflakeFamiliarity with bidding systemsAI-native agentic development styleFormal background in experiment design
Required
We realize applying for jobs can feel daunting at times. Even if you don’t check all the boxes in the job description, we encourage you to apply anyway.
10+ years of experience shipping ML systems to production at web scale
Deep familiarity with empirical methods for showing the value of marketplace interventions, especially online experimentation, A/B testing, and switchback testing
Expert-level proficiency in Python programming, especially ML and data-related Python
Extensive experience with Snowflake, BigQuery, or similar large-scale columnar store databases
Direct experience with advertising technology architectures and subsystems integration
Familiarity with bidding systems, pacing engines, 1st or 2nd price auctions, and 2- or 3-sided marketplace optimization
AI-native agentic development style, including familiarity with adversarial agentic code reviews and agentic math/conceptual reviews
What Can Help Your Application Stand Out:
A graduate degree in a STEM field—an MS or PhD, valued research-focused master’s degrees with hands-on experiments
A history of close collaboration with product stakeholders
A formal background in experiment design, especially experiments in multi-sided marketplaces
Preferred
Graduate degree in a STEM field—MS or PhD
History of close collaboration with product stakeholders
Formal background in experiment design, especially experiments in multi-sided marketplaces