Tech Stack: PyTorch, Python, Pandas

Expert’s cases:

1. During my tenure as a Junior to Mid-level Machine Learning Engineer at SoftConstruct, successfully trained customer Lifetime Value (LTV) and Churn prediction models. Additionally, developed an AI chatbot assistant called "Hoory" using Google's Dialogflow API

2. At Picsart, contributed significantly to the marketing team's success by training a user subscription prediction model. This model effectively identified users with a high likelihood of subscribing to premium accounts, resulting in the lowest cost per subscription rate for advertisement campaigns. Also mentored interns in the development of image clustering algorithms for automatically created Image Collections within the app. Additionally, played a key role in developing a data collection pipeline for internal A/B testing

3. At Podcastle, currently driving advancements in speech technology. Has researched and developed a voice cloning model, creating a pipeline for training human-level, real-time custom text-to-speech using a limited dataset of only 70 sentences from the target speaker's speech. Additionally, have worked on projects involving speaker separation, and internal data annotations tools for ASR model prediction evaluation