Prototypes

DropIt - AI-powered food safety for your pets

Sat, 23 Aug 2025 00:00:00 +0000

As my little pup gets older, she is seeming to develop a taste for my food more than her own.

Every pet caretaker has been there: you’re eating something and your dog gives you that look. I’ve learned how weak I am to that look, while also not learned an exact list of foods my dog cannot have (other than the obvious ones). Did you know some dogs can eat dairy while some breeds are rampant with allergies? So confusing.

DropIt is a simple iOS app that gives you a quick, reliable answer based on your pets. No more typing “My dog is a 2 year-old Cockapoo asking for a bit of my sandwich. Is that safe? My sandwich contains…” while your dog is sitting there salivating.

Type in what you’re considering sharing and get a safety check in seconds. The response tells you whether the food is safe, requires caution, or should be avoided entirely - along with a short explanation and alternative suggestions if needed.

The app uses a chat-style interface that feels natural. You ask about a food the same way you’d text a friend who happens to be a veterinarian. Each response includes a clear safety level, a summary, and any relevant warnings specific to your pet’s breed, age, and weight.

On the backend, Firebase Cloud Functions handle the AI queries through OpenAI’s API, keeping everything server-side so no API keys live on the device. Firestore manages user data, search history, and usage tracking. The app supports Sign in with Apple and anonymous auth for people who just want to try it out without committing to an account.

Building this was a fun exercise in combining SwiftUI with a Firebase backend and an AI layer on top. The core interaction is dead simple, but getting the details right - rate limiting, abuse detection, quota management, cost tracking - turned it into a more interesting engineering problem than expected.

Can my dog eat that? Now there’s an app for it.

NextUp - Track dates in a novel, feature-packed way!

Fri, 12 May 2023 00:00:00 +0000

Almost 700,000 people, along with myself, absolutely love the Year Progress twitter account - made by Filip Hráček. Tracking the progress has been fun, motivating, and at times, frightening. As impactful as it has been, I’ve always needed more. One tweet every three days feels lightweight but it always creeps up on you every time the tweet hits your feed.

I’ve always wanted to have this tracker on my phone, for more dates than just Jan 1 to the following Jan 1, but the date math was presumably going to be tricky to perfect. I spent a couple hours working through the date math in a Swift Playground and essentially built out the core functionality for the app that I’d always wanted. That night, I started building out NextUp using SwiftUI and displaying that data in a fun, animated circular progress view.

After a few days of work, NextUp was working for any date and any user. All dates were persisted using the FileManager built into the iOS app bundle which keeps everything secure for the user. The app will present you Pinned dates for the day, month, and year and past that, you can use NextUp to track any dates and share them out to people in your life!

This app is meant to just work.

Not get in the way or be more than it needs to be.

Track your dates by downloading NextUp on the App Store!

Covid-iOS - Daily count of cases of COVID-19

Wed, 08 Jul 2020 00:00:00 +0000

During the 2020 quarantine, I decided to spend some time learning SwiftUI and a few other technologies that have been on my to-do list for far too long. What better idea than to build an application that spreads awareness of the coronavirus itself? Covid-iOS is an application that displays daily cases of COVID-19 by states/regions in USA. The subsections detail how many of those patients were hospitalized, recovered, passed away and are still testing positive. I was able to use SwiftUI and the MVVM architecture to build out this project relatively quickly and leave room for building future features seamlessly.

To familiarize myself with Continuous Integration and Continuous Delivery (CI/CD) systems within the iOS environment, this project also leveraged GitHub Actions which is run on every Pull Request (PR). This system ensures the project builds successfully and all unit/integration tests pass before the option to merge pull requests becomes available.

The beauty of SwiftUI shined on the project when porting from iOS to including a complementary watchOS target application. After a short amount of time and not many lines of code, the data flow of grabbing and storing API data was incredibly straightforward. Thanks to SwiftUI’s @State and @ObservedObject, the API data was stored universally leaving watchOS to simply grab and display such data. Lastly, the application’s user interface (UI) stays dynamic based on the user’s OS setting on whether they’re using light/dark mode!

All in all using SwiftUI, instead of the industry-standard UIKit, was an effortless and uncomplicated experience. The declarative programming approach of SwiftUI went hand in hand with using the MVVM architecture, resulting in an efficient and versatile codebase.

HackerNews - Browse HackerNews submissions on iOS

Tue, 12 Nov 2019 00:00:00 +0000

Introduced with Xcode 11, I decided to try out the new SwiftUI library to build out the UI of a master-detail application of HackerNews posts. This project pulls Hacker News submissions through the Algolia API and displays them in a SwiftUI List (formerly known as UITableView within UIKit) with their respective HN upvotes. Clicking into any cell will bring up a WKWebView displaying the URL of the post.

SixthBoro - Realtime iOS application for PATH trains

Tue, 06 Aug 2019 00:00:00 +0000

Every city with a reliable public transportation system is seemingly matched with an unbelievably low-quality iOS application to display useful information to commuters. Obviously this proposition stands true for NJ PATH train riders as well. The official PATH iOS application (built with Xamarin) leaves much to be desired whether it be functionality or design. In situations trying to plan train schedules in a hurry, users are often left waiting as the application loads inefficently or left fiddling endlessly trying to make sense of the hard-to-understand user interface.

Version 1.0

After a couple weeks of designing, prototyping, and refactoring my take on a PATH iOS application, I’m happy to showcase the current state of SixthBoro.

Share train times seamlessly by swiping on the train cell.

SixthBoro, named after the common phrase used to describe Jersey City and Hoboken as being New York City’s next borough, is a lightweight and reliable application to seamlessly check train times. The application is packed with iOS features to improve the user experience and provide a quick way to get the information you need.

SixthBoro allows for users to tap between relative and local times seamlessly.

Contrary to the publicly available official PATH app, SixthBoro’s UI aims to be very straightforward, simple, and clean. There is no interest in bogging down the experience by showing you information you likely won’t need on a realtime basis, such as untimely alerts or links to PATH’s website. The goal is to show you realtime train data for you to better plan your schedule with a personable, iOS user experience. Alongside straightforward functionality, SixthBoro displays useful information like train status, train line color, and time until a train arrives to the given station. Lastly, if you happen to pull up SixthBoro at a later time, you can even pull-to-refresh the train data and get a realtime result.

Favorite the stations you frequent.

Commuters can now set favorite stations which they commonly frequent reducing the work of navigating throughout the application just to find desired results. With SixthBoro, you can even set up notifications for specific trains giving commuters the comfort of not having to check back on train times so often. Another simple feature we’ve gotten great feedback on is the ability to switch between relative and local time providing better context on scheduling times.

It has been incredibly fun building an application that could be immensely helpful to any commuter traveling by way of PATH on a regular basis. Feel free to leave me a note if you have any suggestions or comments at sixthboroapp@gmail.com!

Status

The application is now live on the Apple App Store! You can try the most recent version on your iOS devices and hopefully have a great experience in doing so.

AR Posters - Music video posters with Augmented Reality

Tue, 23 Jul 2019 00:00:00 +0000

I’ve been interested in learning how iOS applications can provide engaging Augmented Reality experiences since Apple released the ARKit framework in 2017. Much to my surprise, the framework is backed by comprehensive documentation making it incredibly straightforward for someone with proficient iOS knowledge to build an AR experience. One capability within ARKit I decided to hone in on was the quick image detection. With the capability of detecting if a specified image appears in the camera’s view, I decided to overlay the image with a music video using the rendering engine SceneKit.

Overlaying A Tribe Called Quest’s Low End Theory album cover with the We Got the Jazz music video.

Now that Apple is releasing ARKit 3 with iOS 13, I’m excited to play with the new features such as People Occlusion. This allows for someone to walk in front of or behind an object on the AR plane thus creating an even more immersive experience. With image detection, overlay and people occlusion, the potential for Augmented Reality is vast.

DropDiscuss - Communities for vinyl enthusiasts

Wed, 29 May 2019 00:00:00 +0000

An iOS application that seamlessly lets users create posts and groups related to music album (specifically vinyl) releases. Users can compose messages for a global feed or for their group’s active members. The application is built with iOS, Swift 4.2, and Firebase.

Naval API - Retrieve quotes by Naval Ravikant

Mon, 01 Apr 2019 00:00:00 +0000

This project exposes a JSON API for users to make GET requests and retrieve specific collections of quotes by AngelList co-founder and self-proclaimed philosopher Naval Ravikant.

Landing page for the Naval API

Example JSON output

Qualia - Mindful mood tracking

Sat, 30 Mar 2019 00:00:00 +0000

I took a philosophy course during my senior year at Rutgers called Mind, Machines, and Persons which opened my eyes to the implications that face a society that adopts technology as quickly as it has. A ubiquitous experience is feeling consistently lonely while society is more closely connected online than ever before.

However there are unique individual experiences to the many advancemenets of technology, such as split views on the potential rise of AI guiding society, that can be classified as qualia.

Philosophers often use the term ‘qualia’ to refer to the introspectively accessible, phenomenal aspects of our mental lives.

Mission

This project came about from the notion of capturing the billions of qualia to everyday phenomena. I started off by building the foundation for an iOS application that could store journal entries and displaying them in an engaging manner. Fortunately, Ryan helped take the project to the next level with a great sense of design. We added functionality for logging your mood on a subjective scale, and made the journal entry feature optional for saving a post. Furthermore with the data captured from the mood entry, we built in a line chart that would dynamically update and show the mood progression.

Privacy

It was extremely important for us to make privacy the main focus given the nature of journalling being a very personal experience. Obviously, using a remote database to store user entries were immediately ruled out. Fortunately there was a great solution for data persistence on iOS applications that wasn’t Core Data called Realm. I built in the functionality for Realm to store the entries on the device itself which keeps the entire application experience secure and private (and available offline!).

Although this provides a better experience for users, it is impossible for us to leverage user data to grow the application the way users would want. Without using a remote database, it is impossible for developers to see into user entries (good) but there isn’t a potential to use analytics or machine learning to better understand user behavior (bad). In fact with all entries being stored on the device, all user data can be lost by simply deleting the application or getting a different device. We built in a feature within the application to export entries to a CSV format to combat the issue of losing all user data. However, this is a risk we are more than willing to take with Qualia. Privacy is inherently important with the rise of technology and I’d be happier with slower growth if that meant for a safe, secure, and private user experience.

Status

I’m working on gathering user feedback on the current prototype and implementing features that would be helpful.

Job Explorer - Quickly filter GitHub Jobs API results

Mon, 04 Feb 2019 00:00:00 +0000

A straightforward full-stack application that parses the GitHub Jobs API based off two user-inputted fields of job role and location. The application is built with Node.js, Express, and EJS to take in the API endpoint based off of the user inputs and display the objects that are returned.

FocusLoop - Productivity tracking with watchOS

Wed, 10 Oct 2018 00:00:00 +0000

For a while, I have thought about how great and straightforward Apple Watch’s Activity loops are with displaying your daily physical activity. Imagine there was a similar diagram for your productivity levels.

For optimal productivity, I like to set hourly goals ensuring I have something to consistently deliver to keep me on track. FocusLoop lets users set notification routines (triggered every X hours) asking if you are indeed staying productive. If the user prompts yes, the dialog box disappears. If the user prompts no, it brings you to a table of common distractions to let you log what kept you from your goals.

The paired iOS application then shows (and persists, shout out Core Data) you the log of what held you from your hourly deliverables.

SpellCheck - Using Bloom Filters to recommend fixes to typos

Tue, 25 Apr 2017 00:00:00 +0000

In the Rutgers Computer Science course, Principles of Programming Languages, I was able to learn more about programming language paradigms and how their compilers are built to execute in a specified manner. The coursework delved deep into Functional Programming with Scheme for a good portion of the course, and students were able to build out projects in a completely new approach. For example, we implemented a smart spell check which uses the probabilistic data structure, Bloom Filters, to determine if the word is potentially in a set of properly-spelled words.

CLI Journal - Secure logs for Command Line

Fri, 24 Mar 2017 00:00:00 +0000

This project is a command-line interface application that allows for users to store, read, delete, and search for journal entries. CLI Journal is written with Python, Flask, and Peewee, while the journal entries are stored in SQLite.

Betadrop - Qualitative feedback system

Fri, 21 Oct 2016 00:00:00 +0000

My friend Kyle Ryan and I spent months researching the dearth of quantitative and qualitative feedback prevalent in modern technology platforms. As engineers learning the mobile iOS platform, our mission became:

Build a system that makes it easy for people to get feedback in order to make better decisions.

We did research on feedback pools and eventually delved down the path to solving three major findings: using gestures to provide feedback easier, building a novel feedback interface, and visualizing results that could be understood at a glance.

ConnectFour - Zero sum game for iMessage

Tue, 09 Aug 2016 00:00:00 +0000

At the developer conference WWDC in 2016, Apple had announced the release of an entire App Store comprised of applications solely meant to be shared with another recipient via iMessage. This announcement opened the doors for stickers, games, and posts to be shared between individuals/groups. My friend Amar and I decided to try out building the collaborative zero-zum ConnectFour game solely through the new iMessage API. Before the iOS had publicly launched and the iMessage App Store was eventually unveiled, Amar and I had finished our working copy of the ConnectFour game playable between two iMessage accounts.

Disclaimer: Potential breaking points are iPads and group chats!

Slack Lyrics - Reference Genius song pages on Slack

Fri, 08 Jul 2016 00:00:00 +0000

SlackLyrics is a Node.js application that uses the Genius and Slack API to request a query and return the Genius reults in a Slack message.

Hoot - Online study groups

Sat, 21 Feb 2015 00:00:00 +0000

A couple friends and I came together to learn iOS and ship our first application to the Apple App Store in 2015. Hoot was set up to ease the stress of college classes with organizing all communication into the right pools.

Features

Hoot is a community homework helper and event informer that provides a variety of features:

Ask questions to anyone in your college.
Answer questions from anyone in your college.
Create groups for your classes to ask/answer a specific array of people.
Create events for review sessions/campus wide events.
Inform others that you are going to the event by signing up.
Gain points for performing actions within the app to acquire tokens that can be used to ask more questions.

Prototypes

DropIt - AI-powered food safety for your pets

NextUp - Track dates in a novel, feature-packed way!

Covid-iOS - Daily count of cases of COVID-19

HackerNews - Browse HackerNews submissions on iOS

SixthBoro - Realtime iOS application for PATH trains

Version 1.0

Status

AR Posters - Music video posters with Augmented Reality

DropDiscuss - Communities for vinyl enthusiasts

Naval API - Retrieve quotes by Naval Ravikant

Qualia - Mindful mood tracking

Mission

Privacy

Status

Job Explorer - Quickly filter GitHub Jobs API results

FocusLoop - Productivity tracking with watchOS

SpellCheck - Using Bloom Filters to recommend fixes to typos

CLI Journal - Secure logs for Command Line

Betadrop - Qualitative feedback system

ConnectFour - Zero sum game for iMessage

Slack Lyrics - Reference Genius song pages on Slack

Hoot - Online study groups

Features

Categories