Senior Data Scientist — AI for Mobile Gaming

Senior Data Scientist — AI for Mobile Gaming About the Role We’re looking for a world-class Senior Data Scientist specializing in Applied AI for Mobile Gaming to push the boundaries of how user acquisition, creative production, and gameplay insights are powered by next-gen models. You’ll work across large-scale marketing datasets, gameplay telemetry, and creative assets to architect intelligent systems that help mobile games grow faster and smarter. This is a highly technical, high-impact role—not just building models, but designing the next evolution of voice-over automation, text-to-video generation, LLM interpretability, and video content deconstruction pipelines. You’ll partner closely with product, engineering, and creative teams to take breakthrough research and convert it into production-grade systems used across millions of players and thousands of creatives. If you love cutting-edge AI, thrive in ambiguous problem spaces, and want to shape the future of mobile gaming growth, this role is for you. What You’ll Own - End-to-end model development for marketing, gameplay, and creative pipelines—spanning data ingestion, modeling, deployment, and ongoing optimization. - Advancement of Voice-Over AI technologies including neural TTS, style transfer, emotion modeling, and production-ready audio synthesis. - Innovation in text-to-video systems, designing pipelines that combine LLM prompting, diffusion-based video generation, animation workflows, and creative constraints for mobile ads. - Research & development on LLM decoding and interpretability, enabling better model alignment, controllability, and explainability for creative and marketing use cases. - Deconstruction of video assets for automated tagging, scene breakdown, motion analysis, attention mapping, and actionable creative insights. - Production-scale data science architecture, enabling real-time scoring, predictive performance modeling, and creative optimization for mobile UA campaigns. - Cross-functional leadership, translating complex data science breakthroughs into creative, product, and business roadmap impact. - Continuous exploration of state-of-the-art AI, evaluating new model architectures, agent systems, and multimodal research—and rapidly prototyping what’s most promising. What You Bring - 5–8+ years of experience in data science, applied machine learning, or AI research, ideally within gaming, ad tech, or consumer-scale products. - Expert-level proficiency in LLMs, multimodal models, transformers, diffusion models, and modern AI toolchains. - A deep understanding of ML systems design, including distributed training, model serving, feature engineering at scale, and GPU/accelerator optimization. - Strong experience with large datasets, including marketing data, creative performance data, or telemetry from mobile games. - Proficiency in Python, PyTorch, TensorFlow, JAX, and modern vector databases and embeddings frameworks. - Experience shipping ML models to production environments with real-world performance constraints. - A proven ability to translate ambiguous business needs into concrete, scalable technical solutions. What Will Make You Great at This Role - You think like a systems architect, not just a model builder—seeing how LLMs, audio, video, and game data connect to create end-to-end value. - You are obsessed with the creative side of gaming, and have intuition for what makes mobile ads perform. - You have hands-on experience with video segmentation, multimodal embeddings, audio analysis, or diffusion-based video generation. - You are passionate about LLM interpretability, decoding strategies, token-level insights, and controllability techniques. - You’re energized by greenfield challenges and can move fast from concept → prototype → production with minimal guidance. - You stay relentlessly current on new AI research and can evaluate emerging models with a strong sense for what will materially impact mobile gaming. - You balance deep technical excellence with pragmatism—knowing when to build from scratch, when to leverage open-source models, and when to adapt cutting-edge research.