Projects

Focus

Description:	Distraction free twitter experience using custom filters
Language:	Typescript
Framework:	React Native, Express
Libraries:	Expo, React Query, React Navigation, Recoil
Database:	Firebase

Focus allows users to create custom queries to filter tweets in twitter. It allows users to follow the topics and people they are interested in without any distractions. When a public tweet matching the filter is posted on twitter, it will be picked up by the polling service and will be available for the user in their Alerts screen.

Highlights

Polling service is implemented using Nodejs that scans for filters set by users every 15 minutes (cron) and store the filtered tweets in firestore. The service is hosted in deta which is a free forever cloud provider.

Backend is implemented using Express to handle signup, login, crud operations for filters created by users and fetching tweets that were filtered using polling service. Twitter api requests that are necessary for the application to function are build from scratch with proper error handling, without the use of any wrapper library.

Filters screen is used for interacting with the filters and provides all the details regarding the filter along with operations such as add, edit, pause, and delete. Check out my blog here for a step by step explanation on creating dynamic and custom drawer component.

Alerts screen is designed to in such a way that the UX is exactly same as twitter application. Profile image, profile name, username, tweet’s creation time, the tweet itself are linked to appropriate twitter addresses. Additionally hashtags, links and mentions from tweets are also processed and linked to their appropriate twitter pages.

React query is used for handling network requests made by the application. Some of the additional features implemented include prefetching data using queries during startup, invalidating stale data in foreground and background pages on mutations (modifications in data such as edit, delete), re-fetching and invalidating stale data when the user pulls to refresh supported pages, and handling errors during requests.

Expo is used to handle status bar, fonts, local storage and controlled browser sessions for signup.

Additionally formik is used for creating a controlled form, yup for client side validation, native-base for UI components and dark mode, recoil for state management and react-navigation for implementing bottom tabs navigator.

Description:	Estimate speed at which pedestrians are walking
Language:	Python
Framework:	PyTorch
Libraries:	OpenCV, Numpy
Duration:	Less than 40 hours to build

Projects

Focus

Highlights

Hashnote

Highlights

Portfolio

Highlights

Figma to Webapp

Visualization

Highlights

Pedestrian Speed Estimation

Highlights

Upcoming

Description:	Tag based Notes application
Language:	Typescript
Framework:	Nextjs
Libraries:	React, React Query, Slatejs
Database:	MongoDB

Description:	Personal portfolio and blog
Language:	Javascript
Framework:	Nextjs
Libraries:	React
Database:	Notion, Firebase

Description:	Figma to high fidelity responsive webpage
Language:	Javascript
Framework:	React
Libraries:	Tailwind, Chartjs

Description:	Attempt at data visualization using react
Language:	Javascript
Framework:	React
Libraries:	Redux, Chartjs, Leaflet
Database:	Github
Duration:	Less than 40 hours to build