About the role
Genies is an AI avatar technology company powering the visual and embodiment layer for AI. Through the Genies SDK, their proprietary avatar framework brings AI to life as fully realized AI personas that look, think, behave, and play like living identities. Genies powers AI companion, NPC, and agent use cases across enterprise, consumer, creator, entertainment, and gaming. Genies has raised $200M from Silver Lake, BOND, and Bob Iger. Genies is built on the belief that every website, app, game, brand, and human will have an AI persona. Genies is looking for a Senior Machine Learning Engineer to join our Avatar Technology team, focused on building the next generation of AI-driven animation systems. This role is focused on developing the core machine learning models that power animation generation, including diffusion and multimodal systems. You will work on problems at the intersection of motion, behavior, and generative AI, defining how avatar motion is created and represented. What You’ll Be Doing: Design, train, and deploy machine learning models for animation generation, including text-to-motion, audio-driven facial animation, and full-body performance systems. Explore and implement modern generative architectures, including diffusion models, transformers, and hybrid approaches for motion synthesis. Build multimodal systems that connect inputs such as text, audio, and intent to coherent motion output. Drive the full ML lifecycle, including data processing, training, evaluation, optimization, and deployment. Develop representations for motion, including skeletal animation, facial systems (FACS/blendshapes), and latent motion spaces. Work with large-scale animation datasets (mocap, video, procedural data) and design training pipelines for motion prediction and generation. Define evaluation methods that balance quantitative metrics with perceptual animation quality. Collaborate with Behavior and LLM teams to connect motion systems with higher-level intent and character logic. Optimize models for performance, including real-time or near real-time inference where applicable.. What You Should Have: 7+ years of experience in machine learning, with a focus on generative models, multimodal systems, or related areas. Strong foundation in machine learning fundamentals, including probability, optimization, and deep learning architectures. Hands-on experience with generative models such as diffusion, transformers, VAEs, or GANs. Experience building and training models using frameworks such as PyTorch. Experience working with multimodal data (e.g., text, audio, video, or motion). Experience owning end-to-end ML systems, from research through production deployment. Background in 3D animation systems, motion data, or computer graphics is strongly preferred. Familiarity with model optimization and deployment techniques (e.g., ONNX, real-time inference) is a plus.