Developed Uzbek STT Model: Led a team in the development of a state-of-the-art Uzbek Speech-to-Text (STT) model, achieving a Word Error Rate (WER) of 11%. Built TTS Model: Contributed to the development of a Text-to-Speech (TTS) model, leveraging deep learning techniques to generate high-quality synthetic speech. Optimized Large STT Model: Successfully optimized a large STT model, reducing its size from ~6GB to just ~66MB, while maintaining high accuracy and performance.

• Designed and developed analysis systems to extract information from large scale data.• Developed and evaluated algorithms using Machine Learning and Statistical modelingtechniques to increase performance, quality and accuracy• Experience using TensorFlow and Keras to train and deploy CNNs, LSTMs, and other Sequencemodels

• Managing a team of 5 people• Coordination of timely launch of purchasing projects • Sourcing strategy setting • Budget planning – making sure the targets are achieved