Personalize & store your pics with Captions & Audio.
Sensors, Microcontrollers, a Web-portal and an App to help patients take pills on time.
Plan Bart trips with a simple Web / Mobile app.
Java app to decrypt substitution ciphers.
Computer Science & Engineering
Santa Clara University
Electronics & Communications Engg
Visveswaraya Tech University
Robert Bosch Engg & Business Solutions
Bangalore, India — Sep '11 - May '13
Robert Bosch Automotive Systems Corp
Yokohama, Japan — Jan - Mar '13
Your Trip Pal helps in organizing and planning trips. Provides a simple and elegant itinerary view with all the important trip details. Allows to create and manage trip To-do List, share itinerary with friends when you are done planning. Existing packing templates helps create and manage packing list. Set reminders for planning, packing etc. One can get automatic planning reminder notifications for your trips, create and maintain bucket list of places to visit, along with the best times to visit them and have access to all your past trips at all times.
Key Features include
Google Places API support for adding a trip, adding transport, hotel and event reservation details to the trip.
Location Services, Geolocation & Geofencing API and Camera API support for the "Say Hello" feature.
"Say Hello" to friends: When you reach your trip destination, you will get a reminder to say hello to your friends and family to let them know you have reached safely. You can say hello with a pic and a message in one click.
Alarm Broadcast receivers for notifications.
SQLite for storing user's trip data locally.
Your PA helps track a prioritized task list efficiently. It allows specifying locations for tasks and provides location and time based reminders to attend to them. Supports ordering tasks by priority / task ETA for ease of organizing and scheduling.
Your PA uses Location Services, Geolocation & Geofencing API to provide location based task Notifications.
An App that’s a peppy addition to a camera which lets you take notes as and when you click. Notes could be both Text and Voice and you could also splash your own custom strokes on the pics for a fun personal touch.
Voice Notes feature uses the Media Recorder/Player objects.
Photo Editing tracks Motion to allow custom strokes on the pic and Accelerometer sensor is used to reset edits upon shaking.
Google maps integration is done using SupportMapFragment.
Pill Cabinet Circuit ESP8266 Chip
Uses ESP8266, an Android app & a web portal to smarten the medical cabinet and help manage prescriptions efficiently for self and family members. Provides pill intake & refill reminders and helps monitor adherence and more.
Key Product Features of the Smart Pill Cabinet are:
Timely Reminders : When it is time to take a pill, the Pill cabinet will buzz and also light up an LED. The user will also receive an Android notification listing the pills due then. Inside the cabinet, the bottles with pills due will be illuminated RED.
Adherence Monitoring : Distance ranging sensors monitor the Cabinet open/close action to figure out if pills are taken. We also ask for user confirmation to press a button on the pill bottle after taking the pill. Post confirmation, to prevent repetition of pills we change pill bottle LED color to WHITE. If any pill was missed, the family will be notified in the app.
Distance sensors help measure existing pill quantity in real time. In addition to this and the adherence history/ prescription from server, timely refill notifications are triggered.
Arduino C++ Framework for ESP8266 microcontroller was used to interact with the sensors HCSR04 , trigger time based actions and communicate with the server to implement the various features.
Android App uses the The Async HTTP Client to fetch timely info from the server and trigger notifications. It also provides a frontend for users to login and view their prescriptions.
The server was implemented in JSP and stores / exposes User prescription data used by the Android & Arduino clients. It also provides a frontend for users / doctors to login and view / prescribe medication.
Server reads BART publicly available info and exposes it to clients as Json based REST API. Web & Mobile Clients use this API and allow users to interact and plan trips using BART.
PHP, CURL: The server is written in PhP and fetches data from BART site using CURL and exposes JSON based REST APIs.
Phone Gap: Phone Gap app is used to convert the web app to Android / IOS apps and is made available on mobiles.
Decrypts a Simple Substitution Cipher text. Creates Knowledge Sources with enough knowledge about the english language with respect to the spellings of the words, vowels, consonants, grammar etc, so that it is able to decrypt the input by making a series of intelligent suggestions about the mappings and evaluating them against the sentence database.
Backtracking is used to explore the search space (which is the sentence database). Inverted Indices on words and sentences are used to search the database swiftly and make the decryption faster. The BlackBoard Design Pattern is followed in the implementation.
Bubblewrap: Personal Digital Archiving - Emails.
This was an entry for Personal Digital Archiving Hackathon Stanford 2017. We were a team of 4. Our team won the award for "Most Innovative Project relating to Email Management".
Most artists work with various teams to create art. And lot of content, be it pictures, conversations are shared over emails. There is a huge need to preserve this content and make it easily accessible for people interested. Be it researchers, students or art lovers. This problem is being addressed by Bubblewrap.
Bubble Wrap makes it easy for artists to share content (pictures, conversations) they want the world to know. And provide an organised way for researchers/art lovers to access it and help them get more insight into how the art came into life. Helps preserve the essence of art.
Bubble Wrap provides a People-Centric View of Emails. And allows you to choose subjects you want to share. And after permissions from all people involved has been received, it creates different kinds of views for the material, so as to make it easy for people interested to access it. Artists have the option to share these collections with colleagues, researchers, archivists, or personal contacts.
The people-centric view or the Relationship Graph is generated by using the JSON object that is created based on the header information in the emails. Python, NLTK(Natural Language Toolkit) is used to cleaning, pre-processing email data, extract header information to create JSON objects(which contains all the nodes(people) and the links between them) and extract the email content to generate Extractive Summaries of it. JSON object mainly depicts the number of email subjects each person is involved in and number of shared emails between every 2 individuals.