Purpose of the position
- Identify and solve advanced analytics problems(PTNC) that have the potential to be applied in operational processes to improve the effectiveness of data-based decision making and data analysis data science (DTDL).
- Coordinate the monitoring of the performance of models that have been operated in practice to make decisions on retraining the model if necessary.
- Responsible for the technical assessment and verification of PTNC models developed by other units.
- Support the development of programs and organize training to improve analytical skills and understanding of data for data analysis personnel at operational units in the Bank.
Job description
a) Identify and solve data science (DTDL) problems that have the potential to be applied in operational processes to improve the effectiveness of data-based decision making and data analysis
- Learn the operational processes of units, discuss and come up with ideas about possible problems potential for implementation, assessment of the importance and feasibility of each problem to assess the priority level and allocate resources for implementation accordingly
- Directly participate in the implementation of solving problems that are assessed to have impact and feasibility, ensuring that the model development process complies with regulations and procedures for implementing data science project development
- Coordinate with the Model Operations department in the process of putting the built model into the operating environment.
b) Coordinate to monitor the performance of models that have been operated in practice to make decisions on retraining the model if necessary
- Set warning thresholds and coordinate regular monitoring of PTNC models during operation.
- Coordinate with the model operations unit to perform calibration and maintenance of PTNC models in case of lower performance allowable threshold.
- Propose retraining of the model in case of detecting signs of data drift, model drift
c) Responsible for evaluating and verifying the technical aspects of the results of advanced analysis models developed by other units
- Receive PTNC model products, conduct verification and assessment of the accuracy of PTNC models developed by other units (including PTNC models built by external partners) and the reasonableness in the steps of model construction and decisions made from sample selection, data cleaning, variable construction, algorithm selection and model training
- Provide technical advice and recommendations to improve model performance (if necessary)
d) Improve analytical skills and understanding of data PTDL human resources in the Bank, ensuring the capacity to proactively solve business problems of units/departments based on data analysis
- Build a framework and conduct capacity assessment of existing PTDL human resources in business blocks, develop training programs/roadmaps to improve PTDL human resources capacity in each stage
- Directly evaluate the results of each employee participating in the PTDL human resources training measurement program, take overall responsibility for the output results of the employees participating in the program training program.
