✓ Led the development of a RAG-based Q&A agent with a team of 4 data scientists to support branch staff in answering queries, improving answer accuracy from 46% to 96% using LangChain and Gemini.
✓ Enhanced the Transaction Prediction model, increasing model performance by 5% to exceed 85%, while also covering additional transaction types using Python, PySpark, and statistical techniques.
✓ Developed a referral model for Tangerine to Scotia Wealth Management, contributing to $3.7MM in transferred funds and $19MM in the pipeline using LGBM.
✓ Analyzed recurring money movements of GWM clients using PySpark, Python, Pandas, and Seaborn, helping to identify and mitigate an average of $6.1MM in monthly retail balance attrition risk.
✓ Mentored data science interns by supporting their technical growth, reviewing their work, and encouraging effective team collaboration.

✓ Designed Customer-Value-Index (CVI), Affluence Index, and Engagement Index, which are successfully leveraged for targeted campaigns and incorporated into datasets as input to machine learning models.
✓ Built a Chequing cross-sell model achieving a 1.25% increase in response rate and driving approximately 2,500 new product sales per campaign using Python, XGBoost, Scikit-Learn, BQ, GCP AI Platform.
✓ Implemented 3 end-to-end machine learning pipelines for data collection, preprocessing, and model prediction using Python, Docker, AirFlow, and GCP BQ, achieving a time savings of approximately 30% per run.
✓ Created a customer segmentation model using KMeans clustering and Factor Analysis to group customers by their journey, supporting a client-first approach and enhancing customer primacy.

✓ Engineered an automated large-scale data pipeline to Extract, Transform, and Load (ETL) data from Enterprise Data Warehouse to GIC pricing model using SAS, SQL, DB2, PySpark, Bash Script, and crontab.
✓ Monitored weekly data pipelines to machine learning models, resolved the incurred data issues in the ETL process, and generated 2 new data features that helped to optimize customer clustering models.
✓ Performed Exploratory Data Analysis (EDA) on 7 years of GIC sales dataset to gain insights into customers’ behavior on GIC laddering using Python, Pandas, Matplotlib, and Seaborn.

✓ Led a team of 4-5 software developers and helped them in different phases of the software development cycle.
✓ Interacted with stakeholders to get a high-level understanding of requirements, analyze requirements, and ensure alignment of requirements, application specifications, and database architectural design.
✓ Designed and implemented a scalable service-oriented software architecture for 8+ applications including SSO, Work Order Automation, Inventory Management, etc. using C#, ASP.NET, MVC, SQL, and LINQ.
✓ Developed database entity models and coordinated Data Migration and Extract-Transform-Load (ETL) from legacy systems databases to the afresh data models using SQL Server Integration Services (SSIS).
✓ Maintained the operation of 5+ legacy software applications written in VB6, added new functionalities followed by debug and test to ensure fulfillment of business needs, reliability, and efficiency of software.

✓ Programmed back-end core functionalities of 4+ web-based office automation applications and reports using ASP.NET, C#, ADO.NET, HTML, CSS, JavaScript, BootStrap, and Crystal Reports.
✓ Designed 5 cross-browser compatible websites using ASP.NET, C#, HTML5, CSS, and Javascript.
✓ Implemented database models and entities including tables, views, stored procedures, and functions in MS SQL Server as well as writing SQL queries to aggregate data for reports using SSMS.
✓ Maintained and enhanced 9 websites’ user interfaces continuously in collaboration with 4 other developers.
✓ Generated 10+ RESTful APIs using C# that enabled third-party applications to fetch real-time data.
✓ Instructed 100+ end-users regularly on using applications that aided them in leveraging features efficiently.