K2Neet

K2Neet is an AI study app for students preparing for NEET, India’s entrance exam for medical colleges. A student picks a subject, asks for notes, a quick answer or a summary, and the AI writes it.

The product is the work of a college friend of ours. He built the product and its AI, asked us to take the frontend, and we built the app together. Sending a question to an AI is the easy part. Making the answer worth reading is the frontend’s job: a raw reply arrives as one long string full of symbols, and nobody can study from that on a phone.

We split the work cleanly. The product and its AI stayed with the founder. We owned the Flutter frontend end to end: the interface designed from scratch, state management, caching, the answer parser and the release to Google Play. The partnership was made formal too: we travelled to Surat to sign the agreements in person.

Three things shaped the app. The home screen is the first thing a student sees, so it should feel alive even when nobody touches it. A long history turns into a wall of text, so each saved item needs to be recognisable before it is read. And the raw reply should never reach the screen: every answer gets parsed into clean, structured text first.

A star breathes above the words Start Your Journey while a shimmer crosses it from corner to corner. Each subject has its own colour and icon, and a thin band of light travels around one card at a time, so the screen never looks frozen.

For anyone unsure what to ask, ready-made questions from the chosen subject sit right below the field, one tap away. Long questions are slow to type on a phone, so students can also speak and the app turns their voice into text.

The app never shows the raw reply. It parses every answer, handling bold, italics, code and lists. The question sits in its own bubble, key terms stand out, points get real numbers and the text has room to breathe. When an answer is worth keeping, it can be exported as a PDF.

Everything a student asks is saved under its subject. Each row carries a soft gradient in the colour of its type: purple for summaries, green for Q&A and orange for notes, so they can be told apart before a word is read. One round button starts something new: the list blurs, the three options rise one after another and the plus turns into a close button.

Screens full of small animations need careful state. We used GetX, so a value is marked as observable and only the widgets that read it rebuild. When the coin balance changes, the coin pill updates and the rest of the home screen stays as it is.

Asking an AI means a network round trip, and students often come back to the same answer. So an answer only travels once: the first open fetches it, and every open after that reads from a cache on the phone with no network call at all.

Typography does quiet work as well. Instrument Serif is kept for rare moments such as the app’s name and Start Your Journey. Everything a student reads sits in Satoshi, which stays clear on a dark screen, with its weights doing the ranking.

In October 2025, K2Neet went live on Google Play in beta. The app ships with the living home screen, four subjects with ready-made questions and voice input, formatted answers with PDF export, a history for every subject, and a cache that opens repeat answers without a network call.

The AI writes the answers. We built everything a student touches.