Configuration

This project was bootstrapped with Expo.

In the project directory, you can run:

1. npm install
2. update EXPO_PUBLIC_OPENAI_KEY inside .env file using a valid openai token
3. npm run ios
4. then follow the prompts on screen to test the simulators.

Screenshots

Home Screen

---

Menu Expanded

---

Menu Collapsed

---

Chat Screen

---

Chat Prompt Screen

Test Notes:

Test requirements

• The chat-UI should have text streaming support ✅
• The chat-UI should expand / collapse the list of tasks under the heading of “tasks” in the task pane; ✅
• Adjust to different screen sizes and adapt well for mobile layout ✅
• Streaming using SSE (Server-Sent Events) with chatGPT turbo ✅

State management

No state management tool was required aside from component state to keep all the state manageable.

Style

I'd rather use styled components for the style. However, this is a prototype and time was a restriction right now.

Typescript

typescript was used to help the type safety across the project even so, a few places still missing types.

Errors

I'd classify errors by type eg. User errors and Application. Then severety to provide a better experience to the user.

Performance

I did not see any opportunity for performance improvements to justify the usage of React.memo and useMemo for instance.

Improvements

• Add unit tests
• Remove inline css
• Add a linter

Challenges

Learn how Chat-GPT SSE works then implement in react native was critical during the technical decisions made while developing.