[Team Leadership]- Directed a team of 17 machine learning experts to develop text recognition services for ID documents and financial statements, which improved accuracy, efficiency, and security.- Oversaw the development of an intelligent interactive voice response system for the call center, resulting in improved customer experience and a 75% reduction in wait times.[Team Management]- Fostered a culture of effective teamwork, communication, and collaboration that enabled… Show more [Team Leadership]- Directed a team of 17 machine learning experts to develop text recognition services for ID documents and financial statements, which improved accuracy, efficiency, and security.- Oversaw the development of an intelligent interactive voice response system for the call center, resulting in improved customer experience and a 75% reduction in wait times.[Team Management]- Fostered a culture of effective teamwork, communication, and collaboration that enabled the team to deliver high-quality products on time.- Created a positive and productive team environment, supporting colleagues with career planning and development.[Data Management]- Implemented a standardized codebase and data structure for the team, utilizing MLFlow to create a streamlined model training pipeline. Show less

- Collaborated with business units to automate operations and defined API specifications for system programmers in 2 OCR related projects.- Established model retraining SOP and implemented Kibana alerts to monitor OCR model performance, ensuring a minimum accuracy of 90% with a coverage rate of 95%.- Managed project using agile methodology, enabling efficient problem identification, solution generation, and fostering team spirit during stand-up meetings.

[Machine Learning Engineer]- Implemented an object detection and optical character recognition (OCR) model to automate the credit assessment process and remittance operations, resulting in a reduction of 6 and 32 working hours per day, respectively.[Data Analytics]- Built a logistic regression model to assess the probability of default and loss given default for unsecured personal loans, and integrated it into an online lending service to offer loan limits and lending rates… Show more [Machine Learning Engineer]- Implemented an object detection and optical character recognition (OCR) model to automate the credit assessment process and remittance operations, resulting in a reduction of 6 and 32 working hours per day, respectively.[Data Analytics]- Built a logistic regression model to assess the probability of default and loss given default for unsecured personal loans, and integrated it into an online lending service to offer loan limits and lending rates based on the results.- Leveraged Python for web scraping government open data sources encompassing demographics, electricity usage, taxes, income, and company locations to develop a Tableau dashboard facilitating branch location decision-making.[API Developer]- Implemented adapter design pattern to integrate 5 OCR APIs, creating a controller API with a compatible interface for streamlined system integration and improved efficiency. Show less