Jun 2017 – Nov 2018 · 1 yr 5 mos · Noida Area, India

• Developed new generation REST API's with etag, pagination and CSID support for Adobe Sign integrations and Adobe DC.
• Researched and designed API specs to move web-pages based on AJAX calls to API's.
• Implementation and design specs for adding and exposing various features in Adobe Sign
• Worked with technologies such as Jersey, Spring, Hibernate and Backbone MVC
• Exploratory Projects (Machine Learning)
• Ranked as Best Adobe DC Hackathon Project for the idea “Prediction of completion time of an Adobe Sign agreement” among 100 of different ideas submitted

Jun 2016 – May 2017 · 11 mos · Noida Area, India

• Developed Java and JavaScript (Browser and Node.JS) based SDK's for Adobe Sign API's.
• Developed SDK Samples and Unit Tests using Mocha and Chai framework.
• Developed scalable layered architecture for REST API's which could be extended and maintained easily.
• Worked with Java technologies such as Hibernate, Spring, Jersey, JAX-RS.
• Integration Testing using JUNIT framework.
• Worked with other tools such as Ant, Maven, Grunt, Perforce, GIT, JIRA, CodeCollab, Swagger

May 2015 – Jul 2015 · 2 mos · Bengaluru Area, India

• Project : Contextual Based App Recommendation System for Documents
• Developed Contextual App Recommendation technique which suggest applications (mobile/desktop apps) along with the content to be launched based on the context of document.
• Developed probabilistic approach to suggest apps and map the keywords to these apps.
• Worked with technologies and Knowledge Bases such as Freebase API, ConceptNet, AlchemyAPI, CoreNLP, Mondovo (Query Dataset), Google Play (Dataset)
• Developed working prototype integrated with Document Cloud.

May 2014 – Jul 2014 · 2 mos · Bengaluru Area, India

• Project : Event Detection and Theme Identification in Social Media
• Devised an incremental clustering technique to identify themes from stream of twitter posts collected over a period of time.
• Developed approach to incorporate tweet specific, cluster specific and time based cosine similarity between tweet and cluster in order to perform cluster assignment.
• Developed working prototype i.e. website with data visualisation.

Jul 2015 – May 2016 · 10 mos · Haridwar Area, India

• Project : DiNeR - Building Multilingual Disease-News Repository
• Proposed a multilingual i.e. language independent and low-resource “DiNeR” to build and index Disease-News Repository for collecting news related to diseases that could be extended to all the languages without any significant effort.
• Introduced methods to develop language specific disease news corpus and identify new disease name given the language specific corpus of disease names.
• Developed SVM classifier in R in order to classify disease names and type of news.
• Our filtering technique performs significantly better than baseline-approach both in terms of F-Score(>+5%) and recall(>+50%).