Full-time Data Scientist at EXL Service (EXL Health). Computational data analytics and machine learning to enhance healthcare services, clinical decision-making by identifying high-risk, high-cost members for proactive care management interventions. Working as a Data Scientist with multiple cross-functional teams with numerous exciting projects in healthcare industry for various clients across the USA. My job responsibilities include but are not limited to:- Write code for Exploratory Data Analysis (EDA) using SQL, Pandas, PySpark, Matplotlib, Seaborn- Perform ETL jobs to ensure big data is loaded properly from Azure Databricks, HIVE Data Warehouse using SQL and PySpark- Feature engineering to make sure the data is appropriate for ML models using NumPy, Scikit-learn- Perform correlation analysis and feature selection before the ML analytics- Identify and develop the optimal ML models (by conducting hyper-parameter tuning using AutoML) that outperform existing models in production - Write production-ready code and collaborate with technology team to deploy the model- Query optimization to deliver scalable data solutions both using SQL, Pandas, and PySpark- Write code for performance evaluation (i.e., ROC, AUC Score, Recall, Precision, F1 Score, R2, Average savings/PMPM, Gap prevalence, Condition prevalence etc.) of the existing models with the latest client data to ensure the models are performing appropriately- Develop POC (Proof-of-concept) of ML methodologies from various published manuscripts and implement into production- Gather useful insights to impact business decisions using data transformations, quantitative analysis, causal inference, hypothesis testing, statistical significance test, correlation analysis, feature selection- Creating and maintaining technical document/report for data science or AI models, ensuring that knowledge is shared and the work can be replicated by others.

Worked as a full-time Data Scientist & DL Developer (Acoustic) at MERIDIAN (Marine Environmental Research Infrastructure for Data Integration and Application Network), Dalhousie University, Halifax, Nova Scotia, Canada. My role at MERIDIAN briefly involved the following tasks:- Use Python libraries to pre-process low-frequency audio signals- Create data analysis pipeline to classify marine species (i.e., Orca, bow-head whale, etc.) by using acoustic data- Utilized MERIDIAN's Ketos package to extract spectrograms from audio and then trained Deep Learning models to classify spectrograms- Apply data augmentation techniques such as SpecAug to increase the training sample of the DL model, which lead to performance enhancement of the model- Contribute to the development of the Python-based Ketos package (https://docs.meridian.cs.dal.ca/ketos/)- Maintain test data and annotations from various sources- Analyze key performance indicators (KPIs) for multiple test data sources

- Courses conducted as a GA: Software Engineering, Computer Architecture, Smart Health Informatics, Computer Vision, Computer Networks, C Programming Language- Delivered tutorial lecture class for graduate students on ML/DL

- Conducted literature review of different projects and implemented solutions using state-of-the-art Deep Learning techniques- Worked on several ML/DL-related projects using highly unstructured data in signal processing, time-series data, and medical imaging domains.

- Conducted classes of different programming courses. Also, acted as an academic advisor for the students and a member of the undergraduate thesis defense committee. Supervised more than 10 intern students and 3 project groups- Courses conducted: Java OOP, Data Structures, Algorithm Design, Database Design, Numerical Methods

Worked as a full stack software engineer. Wrote maintainable code for web applications using PHP's Codeigniter framework. Designed and managed MySQL database and developed APIs for mobile applications