JOSIAH TETTEY

Astute and problem solving machine learning/deep learning engineer with extensive knowledge in offering success delivering cutting-edge algorithms, prototypes, and proof of concepts using expertise in pattern recognition, anomaly detection, time series forecasting and others as key contributors to the entire software development life cycle (SDLC). Experienced in Python, data analysis, data visualization, data cleaning and developing predictive modeling. Skilled in C++, Java, SQL, and other front-end technologies like HTML, CSS, JavaScript.

• Developed models for credit card fraud detection while dealing with imbalanced classification data of about 80% positives which birthed the use of specific metric evaluations like F1 score, Precision, among others.
• Implemented a customer segmentation algorithm by employing the K Nearest Neighbors algorithm leading to a 32% increase in service and product recommendations for clients.
• Utilized Python's supervised and unsupervised machine learning models as well as CV (GridSearch) model to identify patterns, develop contextualized feature sets and predict the longevity of NBA players, resulting in valuable insights for teams and stakeholders and contribute to the advancement of data-driven decisionmaking in the sports industry.
• Demonstrated advanced proficiency in deep learning frameworks, including TensorFlow and PyTorch, by leveraging these tools to train and develop complex CNN, RNN, LSTM models for Tesla’s Forex Closing Price predictions.
• Conducted a research project on deep learning techniques and anomaly detection for time series forecasting, using ROC models to evaluate the performance of various approaches.
• Leveraged advanced C++ programming skills to design and implement an organizational binary tree structure, enabling efficient storage and retrieval of detailed worker position data.
• Developed a plastic waste management website to link households with sewage companies utilizing advanced skills in PHP, MySQL, JavaScript.
• Designed and implemented an efficient information retrieval system in Java using the Lucene API.
• Assisted in building a novel neural network model for detecting outliers which worked cross-functionally with the data engineers, data analysts and other departments, even to SMEs.