Experience

Tinnews: A Tinder-Like News Android APP
• Adhered with the Google Component Architectural MVVM Pattern in designing the Instagram Flavor News App.
• Implemented the JetPack navigation component for integrating the bottom bar and page navigation.
• Supported the swipe gestures for liking/disliking the news utilizing Mindork's Place Holder View.
• Employed the use of the Live Data & View Model for developing the Room Database to support a local cache and offline model.
• Pioneered the integration of the Retrofit and Rxjava for generating the latest news data from a RESTFUL endpoint (newsapi.org).

Around: A Cloud and React-based Social Network Front & Back End Development

• Spearheaded the design and launch of a geo-based social network web application with React JS.
• Provided user support in creating and browsing posts as well as searching nearby posts using Ant Design, GeoLocation API and Google Map API for implementing various features.
• Optimized the authentication using token-based registration/login/logout flow with React Router v4 and server-side user authentication with JWT. 
• Handled and deployed posts to Google Cloud (Google Kubernetes Engine) by launching a scalable web service in Go.
• Facilitated geo-location-based search functions to enable users to search nearby posts within a distance (e.g., 200km) using and deploying ElasticSearch to GCE.
• Structured and integrated a face detection model with the Go service using Google Vision API.

JOB+: AWS Based Web Service Development – Job Recommendation Front & Back End Development
• Headed the design of a web page for users to search for job positions using HTML, CSS, JavaScript, snd AJAX.
• Generated favorite records to provide personalized position recommendation.
• Handled HTTP requests through designing 3 Java servlets with RESTful APIs.
• Extracted keywords from position descriptions using the MonkeyLearn API.
• Accomplished in implementing job recommendation by designing content-based recommendation algorithm.
• Maximized performance rate through deployment to Amazon.

Camera Calibration Algorithm:
 • Charted the design and optimization of the camera calibration step of the project, which uses AruCo chessboards to estimate the intrinsic (internal characteristics such as focal length image center) and extrinsic (camera’s position and orientation in the world) parameters of the camera system.
• Conceptualized and conducted the refactoring of the algorithm and C++/OpenCV program to detect, check, and print out the number of detected marker corners of new calibration images and calculate intrinsic and extrinsic calibration results.
• Handled the collection and capturing of 10 sets of images of various objects as data for the Benchmark.

Robot Pose Data Tracking:
• Led the design of a python program that set up the SteamVR system and tracked pose data utilizing HTC VIVE tracker.
• The tracked data could be utilized by a robot arm to simulate the trajectory of human arm movement while collecting ambient, calibration, and structured lighting images for the dataset.

• Oversaw the creation of unit tests that evaluates and tests the functions of a web-based cash flow product using ABAP language.
• Effectively utilized START (with customer code) to design automation framework that integrate multiple testcases.
• Spearheaded the creation and performance of around 30 test cases for a web-based cash flow finance product.

• Piloted the creation of two versions of websites to develop an understanding of customer’s behavior on visiting Gravity Idea’s website (A/B testing).
• Administered the extraction and analysis of the psychological behavior of 500 to 600 Twitter users of Gravity Ideas using spaCy.
• Conducted the design utilizing FUNF and collected data from an Android application of a psychological living lab to understand people’s habitual path to work.