In this role you will advance dynamic pricing and recommender systems for Tripledot’s games, focusing on data understanding, feature design and model improvements. You’ll collaborate with monetization and product teams to plan A/B tests and align ML objectives with business goals. Early months emphasize contributing to the training pipeline and addressing model issues, with increasing ownership of model direction over time. You’ll operate in a fast-paced, cross-functional environment that values impact and scale, using ML to drive revenue and player outcomes.
Compensaciones / Incentivos
- 25 days paid holiday
- Hybrid Working
- Private Medical Cover
- Life & Critical Illness Cover
- Family Forming Support
- Continuous Professional Development
Responsabilidades
- Build and improve training pipelines for dynamic pricing and recommender models, including feature design, training, tuning and offline evaluation of tabular models (neural nets and gradient-boosted trees).
- Monitor live model performance across games and products; investigate data drift and extend models to other products.
- Diagnose missed outcomes across training, business logic and data; implement improvements to restore quality.
- Collaborate with monetization and product teams to link model decisions to revenue and player outcomes; plan A/B tests.
- Identify gaps in the team's understanding of models and propose directional improvements.
Requisitos principales
- Hands-on training and evaluation of tabular models using neural networks or gradient-boosted trees.
- Experience with PyTorch, PyTorch Lightning, TensorFlow, XGBoost, CatBoost or scikit-learn (any framework acceptable).
- SQL proficiency; ability to write efficient queries for relational data manipulation and aggregation.
- Investigative approach to incomplete data and longer-term requirements needing clarification.
- Ability to explain model results to monetization and product partners in revenue/player terms.
- Experience in ad tech, recommender systems or online marketplaces is useful; production APIs or deploying ML models is a plus; Ray framework is a plus.
- Collaborative mindset with cross-functional teams
- Strong analytical and problem-solving abilities
- Effective communication of ML concepts to non-technical stakeholders
- PyTorch
- PyTorch Lightning
- TensorFlow