Data Scientist (Gen AI)

Location: Hybrid/Onsite Team: AI/ML R&D Type: Full-Time Were looking for a highly skilled, hands-on Data Scientist with 410 years of experience in applied AI/ML to join our fast-paced, research-driven team. This role demands deep expertise in transformer architectures along with strong fundamentals in model training, fine-tuning, and optimization. You will work across text, audio, and video modalities , with the opportunity to specialize deeply in one area while maintaining the versatility to experiment across others. If you thrive in a startup-style, high-velocity R&D environment and enjoy taking ownership from architecture to deployment, this role is for you. What You'll Do Model Development & Fine-Tuning Run end-to-end experiments on transformer-based architectures (LLMs, Whisper, diffusion, LoRA, RLHF/SFT, multimodal models). Prototype, benchmark, and push the limits of emerging generative AI techniques. Domain-Specific Applications Audio Lip-sync accuracy Emotional delivery (whispering, shouting, crying) Regional language model support Video Character & scene consistency Quality outputs comparable to Veo3/Sora Motion smoothness & frame coherence Text Extend LLMs to handle regional languages Build domain-adapted, fine-tuned models Evaluation & Optimization Build automated evaluation pipelines for audio, video, and image quality scoring. Optimize trade-offs between speed, quality, and compute efficiency. Cross-Modality Integration Experiment with audiovideo synchronization. Integrate background score generation and text-to-video alignment. Push the boundaries of multimodal generative systems. Research & Experimentation Stay ahead of rapidly evolving AI models, tools, and techniques. Test architectural variations and scaling strategies for production-ready systems. Ownership & Execution Take complete ownership of projects from idea prototype deployment. Demonstrate strong problem-solving, accountability, and first-principles thinking. What You'll Need Experience 410 years in applied ML/Data Science with significant generative AI experience. Core Fundamentals Deep understanding of transformer architectures, training dynamics, and model optimization. Experience with LLMs, diffusion models, and multimodal architectures. Modality Expertise Depth in at least one modality (text, audio, or video) with end-to-end project delivery. Technical Skills Strong coding skills in Python . Expertise with frameworks like PyTorch or TensorFlow . Experience with deployment frameworks (e.g., FastAPI , model servers). Evaluation Experience Ability to design and implement automated evaluation systems for generative outputs. Adaptability Comfort with rapid experimentation and learning new tools, models, and techniques. Location: Hybrid/Onsite Team: AI/ML R&D Type: Full-Time Were looking for a highly skilled, hands-on Data Scientist with 410 years of experience in applied AI/ML to join our fast-paced, research-driven team. This role demands deep expertise in transformer architectures along with strong fundamentals in model training, fine-tuning, and optimization. You will work across text, audio, and video modalities , with the opportunity to specialize deeply in one area while maintaining the versatility to experiment across others. If you thrive in a startup-style, high-velocity R&D environment and enjoy taking ownership from architecture to deployment, this role is for you. What You'll Do Model Development & Fine-Tuning Run end-to-end experiments on transformer-based architectures (LLMs, Whisper, diffusion, LoRA, RLHF/SFT, multimodal models). Prototype, benchmark, and push the limits of emerging generative AI techniques. Domain-Specific Applications Audio Lip-sync accuracy Emotional delivery (whispering, shouting, crying) Regional language model support Video Character & scene consistency Quality outputs comparable to Veo3/Sora Motion smoothness & frame coherence Text Extend LLMs to handle regional languages Build domain-adapted, fine-tuned models Evaluation & Optimization Build automated evaluation pipelines for audio, video, and image quality scoring. Optimize trade-offs between speed, quality, and compute efficiency. Cross-Modality Integration Experiment with audiovideo synchronization. Integrate background score generation and text-to-video alignment. Push the boundaries of multimodal generative systems. Research & Experimentation Stay ahead of rapidly evolving AI models, tools, and techniques. Test architectural variations and scaling strategies for production-ready systems. Ownership & Execution Take complete ownership of projects from idea prototype deployment. Demonstrate strong problem-solving, accountability, and first-principles thinking. What You'll Need Experience 410 years in applied ML/Data Science with significant generative AI experience. Core Fundamentals Deep understanding of transformer architectures, training dynamics, and model optim