Staff / Principal AI Researcher, Inworld AI

• PhD or equivalent experience in Computer Science, Machine Learning, Artificial Intelligence, or a related technical field
• Strong background in machine learning research, demonstrated by publications or significant contributions to the field
• Experience in training large models and working with state-of-the-art approaches in machine learning
• Proficiency with ML frameworks such as PyTorch, TensorFlow, or JAX
• Knowledge of advanced AI architectures, including transformers, diffusion models, and reinforcement learning techniques
• Programming skills in Python or similar languages
• Familiarity with pre-training, fine-tuning and evaluating LLMs such as LLaMA, Mistral, Qwen, etc., is a significant plus

Desirable

• We are seeking Staff and Principal level AI Researchers with extensive experience in researching, designing, and implementing machine learning models and algorithms
• You will be at the forefront of advancing our AI capabilities by developing state-of-the-art models and contributing to cutting-edge research in artificial intelligence
• Your work will involve exploring novel ideas, training large-scale models, and collaborating closely with other researchers and engineers to push the boundaries of what's possible in AI
• Conduct cutting-edge research to advance our AI models and capabilities
• Develop and implement novel machine learning techniques, experimenting with new architectures and algorithms
• Train and evaluate large-scale ML models, optimizing for quality, latency and cost
• Collaborate with cross-functional teams, sharing insights and integrating research findings into our products
• Communicate results internally and publicly, contributing to the broader AI community
• Mentor junior researchers and engineers, fostering a culture of learning and collaboration
• Exploring the impact of new attention mechanisms within transformer architectures to improve model quality and latency
• Investigating methods for 3D animation synthesis by performing ablation study to thoroughly compare diffusion vs flow-matching based approaches
• Developing algorithms for on-device AI, optimizing transformer-based models for specific hardware constraints
• Researching and implementing techniques for creating new AI voices, enabling developers to generate and use them at runtime