kdabhadk

As a Staff Data Scientist at KETOS, I use machine learning and statistics to model water quality parameters. I design, maintain and deploy anomaly detection workflows for high-frequency IoT sensor readings and engineer data science workflows to analyze large datasets. I also design statistical experiments, draft comprehensive reports, and facilitate data accessibility through the KETOS Data Science portal. As a mentor, I guide interns through various projects. I was responsible for designing an interactive Power BI report and creating unique metrics and risk scores. Lastly, I managed the ingestion of public water data points into the KETOS data lake, created insightful dashboards, and led the design of an interactive pipe network visualization tool.

As a data scientist, I developed an NLP system and Power BI dashboards for customer insight extraction, built a proactive recommendation engine, and modeled the behavior of 200 million website visitors for marketing performance evaluation. Additionally, I contributed to improving data science and data engineering technical screens.

I analyzed over six years of reactor temperature data to model and predict the lifetime of industrial chemical reactors. I implemented ARIMA and Keras-based neural networks for time-series modeling and temperature forecasting, successfully diminishing the time-averaged error from 5% to 2.5% on the test data set.