Full-stack developer and overseer of business operations at Genda, with a focus on leveraging technologies such as Large Language Models, Transformers, Machine Learning, and Artificial Intelligence to optimize teaching and learning in the post-secondary space. Incorporating these tools into the custom-built Learning Management System (LMS) and conducting ongoing technological research to improve existing models and enhance user experiences are core responsibilities.

Retrieved thousands U.S. Securities and Exchange Commission filings (10K and 10Q) for various companies to investigate potential correlations between Management Disclosure and Earnings Surprise. Cleaned the data and then utilized NLTK for text tokenization. Employing models and techniques like Bidirectional Encoder Representations from Transformers (BERT) and Large Language Models (LLM), to generate sentence embeddings to feed into a hybrid model. Then fine-tuned BERT to enhance the probability distribution on financial documents. Additionally, developed algorithms and assigned weights to refine the probability distribution. This process culminated in the model providing insights on expected earnings surprise based on the analyzed data.

Collected and organized large amounts of financial data, mainly focusing on stocks, and conducting thorough time series analysis. Using statistical methods like z-score calculations to decipher trends and patterns in the data. Combined data patterns to come up with a composite rank estimate.