Machine Learning Engineer

Who are we?

Tripledot Studios is one of the largest independent mobile games companies in the world.

We are a multi-award-winning organisation, with a global 2,500+ strong team across 12 studios.

Our expanded portfolio includes some of the biggest titles in mobile gaming, collectively reaching top chart positions around the world and engaging over 25 million daily active users.

Tripledot’s guiding principle is that when people love what they do, what they do will be loved by others.

We’re building a company we’re proud of. One filled with driven, incredibly smart and detail-orientated people, who LOVE making games.

Our ambition is to be the most successful games company in the world, and we’re just getting started.

Role Overview

Join Tripledot Studios as a Senior Machine Learning Engineer working on dynamic pricing and recommender systems use cases in our games. The work involves tabular machine learning, including neural networks and approaches such as gradient-boosted decision trees. You'll focus on understanding the data, developing features and improving the models, rather than primarily owning production deployment.

You'll work toward defined near-term ML targetds, collaborating with the monetization team and game product teams to plan A/B tests. Over the longer term, you'll work with those teams to translate product and monetization requirements into ML objectives.

You'll start by learning the project and its business goals. Within the first three months, we'd expect you to contribute to the training pipeline and investigate features and model issues. By around six months, the aim is for you to take a more proactive role in setting the model's direction.

Key Responsibilities

• Build and improve training pipelines for dynamic pricing and recommender system models, from feature and label design through training, tuning and offline evaluation of tabular models such as neural networks and gradient-boosted trees.
• Monitor model performance once models are live across games and products. Investigate data and concept drift, including shifts tied to new titles, client versions or user behaviour, and make models easier to extend to another product.
• Diagnose missed model outcomes across the training pipeline, business logic and underlying data, then work through the improvements needed to restore model quality.
• Work with monetization and product colleagues to connect model decisions to revenue and player outcomes, and help plan the A/B tests that inform what ships.
• As you get up to speed, identify gaps in the team's understanding of the models and propose improvements to their direction.

Required Skills, Knowledge and Expertise

• 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 could all be relevant; no single framework is required.