Machine Learning Engineer

TTEK Inc. focuses on creating and implementing innovative technologies to enhance the efficiency of customs and border processing.

• Used Large Language models with advanced prompting techniques for text to sql. Created one LLM agent in order to do entity extraction. Then used output of that agent in SQL agent in order to create an sql query.
• Utilized advanced Prompting techniques with LLMs (gpt4-o and Claude) to extract valuable insights from diverse data sources in order to create risk rules in customs.
• Created an LLM Agent using an LLM (gpt4-o) together with serper api in order to predict the NAICS code for a given company name.
• Used LLMs for price prediction.
• Designed a dual-head classification model leveraging the BERT LLM architecture for precise categorization of products into Harmonized System (HS) codes.
• Built a Named Entity Recognition (NER) model to extract product names from descriptions, using BERT based LLM for token classification.

• Engaged in research and development activities.
• Developed a REST server for the Autocorrect module, providing seamless and efficient functionality.
• Leveraged OPTUNA for automatic hyperparameter tuning, optimizing the performance of models and algorithms.
• Employed explainable AI techniques, including tools like Captum, Lime, and SHAP, to provide interpretable insights into model predictions and decision-making processes.
• Conducted fine-tuning of BERT models for intent classification tasks, refining their performance on specific contextual understanding tasks.
• Implemented a PyTorch-based DIET classifier from scratch, enabling broad intent recognition and classification.
• Conducted experiments with RASA's open-source conversational AI framework, exploring its capabilities and potential applications.
• Explored end-to-end dialog models using Language Models (LLMs), investigating their effectiveness and performance in conversational AI tasks.

Voiceloft, an Azerbaijan-based platform that promotes the development of speech recognition technologies by collecting voice data.

• Fully responsible for the ML projects;

• Working on ASR model and leading 2 interns on NLP tasks;

• Fine-tuned Wav2Vec2 baseline model with our dataset.

• Worked on R&D and Data Analysis;
• Applied Sentiment Analysis Model from start to deployment, using Python, PyTorch, scikit-learn, Docker, and Flask.

• Held complete responsibility for overseeing and managing the entire lifecycle of ML projects.
• Successfully implemented a state-of-the-art Azerbaijani Speech-to-Text model using the Kaldi toolkit and Python.
• Developed an Azerbaijani Text-to-Speech model using the Tacotron2 and parallel vegan model for synthesis in PyTorch.

• Fully responsible for the ML projects and crawlers;

• Applied Generic Crawler to get data from any news page, even if they change their design;

• Built the ETL process, getting news data from 200 different Azerbaijani news channels;

• Applied Similarity Detection, Text Categorization, NER, Sentiment;

• Analysis, Keyword Extraction, Summarization to those data;

• Finished Projects: aqreqator.az.