► In less than a year at ReachMobi I have contributed towards the development of 4 applications which are live on the Google Play Store and have a user base greater than 5 million (Messenger Home, Messenger Lite, Weather Home and iTheme).

► Leading the programming tasks / efforts for all the 4 products

► Built features like search, customization, attachments, local databases and conducted A/B tests which helped improve user experience and user retention on the messenger apps by 8%.

► Incorporated and improved AdMob, AppOpen, Mopub, Native and Keyword based ad placements to improve on revenue streams. Monitored Firebase Crashlytics for bugs and helped fix more than 50 crashes, driving crash free users up to 99.2% in the products and company wide while collaborating with people from different teams.

► Booted up and launched a new product iTheme, an iOS themed app which gives a user the best of both (iOS and Android) inside a week.

► Made constant improvements on the apps, eg. notifications, device specific configurations, UI changes, resource management, tracking, instrumentation, remote configs, etc.

► Helped in forming new ideas like building widgets, persistent notifications, rewarded ads, walkthrough changes, etc to drive up default rate and thus daily active usage for the apps.

► Worked with Redshift to query data and exporting it to Excel for creating plots for retention, default rate, walkthrough completion, click through rate, a/b test results, etc.

► Completed most of the deliverables before the deadline and made weekly releases along with documenting the same on Slite.

► Helped resolve compliance-based issues company wide by coming up with smart solutions to tackle the same

► Technologies, Libraries and Tools Stack - <Android, Java/Kotlin, RxJava, ContentProvider, Google Play Billing, Subscriptions and Managed Products, SQLiteDB, Leaflet.js, Location Services, Glide, Tenor API, Google Firebase, Redshift, Excel, GitLab./>

► Worked on two applications for the duration I was at the company. [1] An internal contact management and time-off request application. [2] A survey tool application to be used for conducting surveys and for on-boarding / hiring new employees.

► Integrated Angular and researched (DAR Analysis) on libraries for incorporating analytics on the frontend.

► Introduced Ngx-Charts (D3) to build clean and minimalistic charts. Modularized code, manipulated API data, and created domain entities on the frontend while making relevant changes in the backend to form chart data.

► Charts for counters by the time of submission, submission history, time spent on each question scatter plot, pie/donut/bar optional charts (expandables).

► Routing for segregating survey level and question-level data with tabs.

► Dynamic data-table from the responses API data with sorting, filtering.

► Introduced the concept of filter chips to link drill-down based filtering on question level charts with the data-table.

► Invisible ReCaptcha for having survey level security.

► Resolved open issues on the Ngx-Charts library (text on chart slices, making chart slices clickable with an indication).

► Introduced MongoBee migration scripting.

► Worked on identifying regression tests to be performed.

► Feature for color-coding email statuses sent out to multiple supervisors by seen/unseen color codes, priority, and date.

► Implemented automation tests, unit tests resulting in code coverage of at least 95%.

► Found and resolved bugs in the frontend as well as backend, ranging from minor to critical severity, which greatly improved application performance and user experience for both applications.

► Created and delivered presentations for end of sprint demos featuring all the work done in a sprint for both applications.

► Technologies, Libraries and Tools Stack - < angular 9, typeScript 2, java / kotlin, spring boot, spring data and jpa, mysql / mongodb, jira, junit 5, selenium 3, jenkins, aws ec2, ngx-charts (d3), agile s/w dev, bitbucket/>.

► Creator of IMClassify, a platform for different expertise level individuals or organizations, to develop entire computer vision life cycles for performing tasks related to object detection.

► The product is a platform used to train images for "any object’s" detection.

► A PaaS for data collection, pre-processing, training, testing, and finally deployment using your browser directly.

► Although I worked on the project as a co-op, I was the lead developer for the project.

► Researched business requirements and designed wireframes to come up with a clean and minimal UI.

► Configured the scalable cloud infrastructure and darkflow training environment on AWS.

► Set up the Apache PHP environment for hosting the REST services.

► Configured the relational database environment, created the data model for the application, and wrote scripts to automate the onboarding process for new developers.

► Created REST API services (PHP) to perform CRUD operations on the database. Developed scripts to interact with the deep learning model, to create machine learning models on the go with asynchronous monitoring and test them at various checkpoints.

► Created scripts for model and parameter configuration, data collection, training, testing, model evaluation, and performing RAD for models that had reached near completion stage.

► Developed the React.js client application consisting of the dashboard that consumes the REST APIs to help users navigate through different sections of the training process & cycle.

► Used visual analytics and charts to help the user understand the training process and metrics in an easier way.

► Modified the VGG Annotator open-source tool to fit the platform needs and integrated the same with the REST services. The tool had capabilities to annotate objects using labels predecided by the user.

► Technologies, Libraries and Tools Stack - <react.js, php7, python3, mysql, darkflow, darknet, vgg annotations, ant design, wieldy, gitHub, aws ec2, route 53, trello, vanilla js, html5/css, jquery/>.

► To know more about my work here please visit My Experience @ Champs Software, Inc.

► Chattodo is a product where one can gain insightful knowledge on the basis of accomplishment of tasks with just some people collaborating over chats towards completing a project.

► We at Chattodo are trying to make this possible by building an environment to observe conversations around tasks discussed and marked as done.

► We will try to recognize knowledge that must have been earned in the act of completing those tasks and also assign a metric to it so that it can help for the purpose of assessment and thus candidate evaluation and validation.

► In a gist, Chattodo is a chat-first productivity app, which is designed with small teams in mind. In the age of "move fast, fail fast", it is important to get things done. On Chattodo, users typically chat with each other just like any other messaging platform, but they see a lot more happening than just chatting.

► Available on the Play Store, launched through Product Hunt with the story narrated on Medium. (Chattodo)

► Technology Stack: <java for android, xml, firebase realtime db, cloud functions, php, firebase cloud messaging, firebase remote configs, heroku for other deployments />

► Working at the Department of Development and Alumni Relations (DAR), RIT.

► Working on the management and maintenance of the Drupal web environment for the "Alumni" and "Giving" sites at DAR.

► Successfully deployed 10+ modules, views and 15+ pages from the staging to the production sites in my tenure here as a student worker.

► Modules built include ➞ iModules Rest API ➞ iModules Analytics Client for displaying email statistics obtained from the Rest API. (Chart.js) ➞ Accordion in Accordion ➞ Custom Bootstrap Column Module ➞ Chapters and Groups ➞ FAQ Module ➞ Stories of Impact ➞ Tiger's Tale ➞ Web Forms Module

► In addition to this worked on a series of pages using the above modules, existing modules and deployed them too.

► Technologies, Libraries and Tools Stack: <php, twig, drupal, scss, js, jquery, chart.js, mysql, asana/>

► During my short tenure here, I worked on developing a couple of modules in the trade processing system which accepted trades from restful services in various formats (like FPML) and converted them all to a unified format to make further amendments on them.

► Technology Stack: <java, mule esb, sql, xml, json />

► Ranked first at the yearly "Flow Monster Hackathon" in the Global Technology Analyst Program for incoming technology analysts.

► While working at KeyQual I built a couple of retrieval based bots for the company. NetXit is a chatbot for an online work exchange platform for connecting B2B and B2C. Keyhire chatbot targets the social hiring segment with functions such as Job Search, Applications, Evaluations and Corporate - Candidate Reverse Connect.

► Stack: <tensorflow, nltk, dual encoder lstm, numpy, pandas,the ubuntu dialog corpus for conversational needs and a redis database in the background for retrieving data related to the specific chatbot />

Also, developed a responsive job portal which had the following functionalities: Mobile Compatibility, Authentication, SSL Certification, Runtime Notifications, Google Maps API Integration, Pagination, Filters, Social Sharing. ► Extracted data (in structured form) from web portals in huge numbers and managed storage of the same in the database at runtime. One of the key features in the making was the creation of a sub job portal dynamically just at the press of a button along with a completely responsive admin dashboard for the new registration.

► Stack: <python, php, html, css, javascript, jquery, ajax, angular, json, beautifulsoup, sql />

► Analyzed data on automotive car components and developed a predictive model for survival and failure of the same. The technique used for modeling failure was Weibull; and Kaplan Meier, Nelson Aalen were the methods used for analyzing survival of the components.

► In addition to the above, also analyzed sales data (Bolero sales in Chennai) and predicted sales forecast of upcoming 6 months for the same using the ARIMA Model.

► Technology Stack: <r, r-shiny />

► Paper I: International Journal of Science Technology & Engineering | Volume 4, Issue 5.

► Paper II: International Journal of Advance Research in Computer Science | Volume 8, No. 9, November-December 2017

► Developed an android application to manage an optimized solution for local logistics and transportation needs for both businesses and consumers in the urban areas.

► Technology Stack: <java for android, node.js, socket.io, postgresql, google maps api, tookan api, json, aws />

► Developed an Android Client for Google Adwords which automates the process of keyword bidding and managing campaigns.

► Stack: <java for android, google adwords api, oauth 2.0., php />

► Developed the internal website and inquiry portal for Air India Engineering Services Ltd.

► Stack: <html, css, javascript, jquery, asp.net, sql />