Full-time. Remote work possible.
Mandatory Requirements
- Knowledge of neural network architectures (Transformer, CNN, RNN, GAN, Diffusion);
- Proficiency in training models from scratch and fine-tuning existing solutions;
- Experience with PyTorch, TensorFlow, JAX frameworks;
- Skills in writing code for inference and deploying models in a production environment;
- Understanding of optimization principles: gradient descent, loss functions, regularization;
- A set of ML engineer competencies with a focus on independent algorithm implementation;
- Solid command of Python and C++ programming languages;
- Ability to implement individual model components from scratch: for example, an attention mechanism or a separate layer;
- Deep understanding of the internal logic of model operation (not limited to using ready-made methods like model.predict()).
Project Tasks
- Integration of trained models into the company's closed ecosystem (on private servers), ensuring their autonomous operation and protection from external influences;
- Adaptation and optimization of AI algorithms for operation on protected mobile devices - considering limited computational resources and the need to minimize power consumption;
- Deployment of trained models within the company's closed ecosystem (on private servers), maintaining their autonomous operation, and implementing protection measures against external influences.
