KazakhstanVictory-Technology
Computer Vision Research Engineer
2 нед. назад
800k–1000k USD / monthKazakhstanCISMiddleOnsite
pytorchtensorflowcnnrnntransformersdockerrest apimlops
Role involves designing and training advanced multimodal machine learning models for LegalTech projects, focusing on emotion and psychophysiological state detection from multimedia data.
About the company
- Victory-Technology
About the product
- LegalTech projects
Responsibilities
- Detecting emotions and psychophysiological states using advanced multimodal architectures
- Designing and training CNN, RNN, Transformers, Audio/Video Fusion, and LLM/RAG architectures
- Building pipelines for extracting landmarks, audio embeddings, and spectrograms
- Synchronizing audio/video, performing cross-validation, and implementing Explainable AI techniques
- Optimizing models, applying MLOps, packaging in Docker, and integrating via API
Requirements
- 2+ years of experience in ML (CV / Audio / NLP)
- Proficiency in PyTorch and TensorFlow
- Experience with Vision/Audio Transformers and Multimodal Fusion
- Familiarity with RAG/LLM architectures
- Knowledge of Docker, REST API, and MLOps best practices