➢ Identifying business questions and testing hypotheses using research & jurisdiction scans. ➢ Involvement in end-to-end product development from problem definition, and design to deployment. ➢ Propose predictive analytic capabilities in-house through research, development, and prototyping. ➢ Extracting, cleaning and combining data from various sources like public API’s and Corporate Data Warehouse using SQL queries and Python scripts as required.➢ Presented insights on Municipalities building energy and emission trends by following visualization best practices through the dashboard using Dash – Web programming framework. ➢ Built a time series model to predict daily solar prediction with an accuracy of 93.5% from 5 min interval micro-inverter data points. ➢ Built Citizen Twitter analytics app using Dash from data collected from twint, Twitter data scraper, to understand citizen's views on HRM strategies related to Climate change through storytelling with the incorporation of Natural language processing techniques like sentiment analysis, word cloud, and topic modeling. ➢ Research and experimentation with different Machine learning and Natural Language Processing techniques to categorize and understand the frequently used search terms by citizens on the Halifax website through data collected by Google Analytics API. ➢ Research and development of chatbot incorporating the insights from the Google Analytics data for the most frequent search terms on the Halifax website. ➢ Early testing user for HRM’s Azure Cloud Services by using docker containers for Dash App deployment and Azure blob for unstructured data storage. ➢ Communicating the result to technical and non-technical audiences.

Built a Full-Stack Data Science product that can predict a used vehicle’s price as a pointestimate as well as a range based on its characteristics.➢ Worked with a scraped dataset from Kaggle containing “458,213 raw data points”.➢ Built Gradient Boosting Tree-based model with a point estimation accuracy of 89.52%and an MAE 2,527.49 dollars.➢ Implemented Quantile model Prediction for the vehicle price range.➢ Built Serverless RESTful API for serving the model with AWS using Zappa.➢ Used GitHub Actions for automating CI/CD pipeline and pytest for a test suite.➢ Collaborating through a mentorship with a Senior Machine Learning Engineer atDTN.

Built multimedia search engine using AWS Serverless architecture.➢ Developed front-end as well as back-end with serverless architecture and performed quality assurance.➢ Built a microservices with lambda to execute the pipeline for text extraction from different multimedia and document files using AWS Machine Learning services (Textract, Transcription, Transcribe) which uses Comprehend & Elasticsearch to index the documents stored in S3 Bucket. ➢ Tested the whole system with around 1GB of data consisting of different documents (images, docs, pdfs) and multimedia files (video & audio).

Collaborate with other domain developers to analyze, design, and implement the software products according to client specifications.➢ Worked on 5 different major projects teaming up in a group consisting of 3-4 other domain experts depending on requirements.➢ Built Android Mobile Applications using Fragment API, RESTful Web Services, and other third-party libraries using Java MVC.➢ Code review, testing, and documentation.