Loora is on a mission to revolutionize education and break down language barriers with its cutting-edge AI, building the first-ever personal AI English tutor. Loora offers its users an AI tutor that is always available to talk about whatever they want, give immediate feedback on English skills, and guide them on their journey to fluency.

Your role:

We’re looking for an AI Research Intern to join our R&D team and work on video generation models project for our real-time speaking avatars. These avatars are a key part of how users interact with Loora’s AI tutor. You’ll focus on training and optimizing generative models that produce realistic facial video synchronized with speech. You’ll also work on distilling large models to run efficiently on-device, ensuring smooth performance in real-world usage.

Key Responsibilities:

• Train and fine-tune text-to-video or speech-to-video models for realistic avatar animation
• Explore recent methods in diffusion and transformer-based video generation
• Apply model distillation, quantization, and other optimization techniques for on-device performance
• Work closely with the product, engineering and AI teams to bring models into production

Requirements:

• Currently pursuing a MSc or PhD in Computer Science, Electrical Engineering, or related field
• Experience with deep learning and generative models (for video, image, or audio)
• Proficiency in PyTorch or TensorFlow
• Familiarity with recent work in video generation and current SoTA models
• Bonus: experience with on-device ML, model compression, or real-time inference

Even better if you have:

• Experience working with multi-modal models (e.g., speech + video)