Worked on User Interface stories, bugs and unit tests using React and other related technologies.

Integrated backend API with the frontend.

Built data visualization dashboards as per business requirements using technologies like echarts.

Analyzed the defects raised by the QA team to fix and implement the changes in QA.

Built and implemented common React components.

Worked in an architecture that collects data from multiple upstream resources to parse and send downstream for storage and visualization. Used Java,REST,Spring Boot and Oracle DB for Development.

Worked on 15 end to end Machine Learning projects and successfully deployed them in production.

Designed and developed an API to be used in an existing email spam classifier system using Python. This included extracting out and extending the classification logic from the existing monolithic classification system and providing this API instead, making it much more extensible.

Worked on NLP domain such as text classification, text analysis, classification, keyword finding, entity finding.

Participated in planning and design discussions of features with teams and also designed a few POC’s(Proof of Concepts) for clients.

Developed a gaming portal where a customer can login and select games they want from the wide range of collections and make payments.

Participated in all phases of SDLC while working in an agile environment by coding in JavaEE applications using the spring MVC framework.

Developed an Ecommerce webpage (Front-end) where customers can choose a delivery model and make their payments.

Conceived and built optimized landing pages in HTML and CSS, including AEM integration and cross browser compatibility, to support marketing.

Worked in agile environments and used project management tools like Jira and Git/GitHub.

Implemented data mining and data integration techniques. Analyzed data for actionable insights

Cleaned and validated the data to remove outliers.

Removed 25 % redundant data which helped the model to get 15 % more accuracy than the previous model.

Extracted insights from time series data and was able to achieve a low mean squared error of 2 %.