• Made a research on machine learning for sound, music, MIDI generation.• Performed lots of experiments with SOTA models for music generation, MIDI generation and recognition, style transfer, sound synthesis.• Collected data, developed and trained GANs, transformer models with PyTorch for the music GenAI.• Collected data and trained TTS models FastPitch, FastSpeech, WaveGlow for Speech Generation to increase thenumber of speakers and increase the quality of the generated speech• Designed and developed application for high-end music generation with VST plugins• Researched Machine Learning and signal processing tools to enhance the quality of generated sound applied to sound quality enhancement.• Developed speech processing pipelines for the dubbing service with Whisper and Spleeter transformers• Developed the sentence alteration service to match the lengths of original and translated speech

• Researched Deep Learning and Machine Learning approaches applied to the radio signals classification.• Developed tools for the data labeling, data augmentation.• Applied transfer learning to computer vision models ResNet, Inception, VGG.• Based on the data analysis built custom CNN based classifier which overcome other solutions and achieved 80% of precision and recall for 4 of 5 classes on classification problem.

• Building RESTful APIs with Python• Worked with Django’s ORM database interactions