React Native - Full Stack Developer

Company Description

We are an Edtech company - Operating from Chennai.

Profile : Looking for an enthusiastic - React Native / Python / MongoDB developer - Fresher / Experienced - Work from office, Choolaimedu, Chennai.

CTC : 5-6 LPA

Complete the below task - We shall connect with you -

Once done - please mail us at

The above is a short demo video of the Screening PathoAI mobile app. Your assignment is to semi-vibe code the same experience end-to-end in React Native with a Python plus MongoDB backend. Any Gemini or OpenAI model may power inference, but all traffic must go through your backend.

Treat this as a take-home build that shows you can own a small, production-grade product slice independently. - You will need to share a video recording of how you built this - Should leverage at least one of the AI Code assistants like Cursor / Claude code / Codex / CoPilot / or any other such tool. The prompts and the way you solve the problem is as important, if not more important than the end demo

Ground Rules Checklist

☐ No authentication layer; focus entirely on the core experience.

☐ Persistent memory of runs, sessions, and feedback must survive app restarts.

☐ Client must be React Native (Strictly do not use Expo).

☐ Backend must be Python (FastAPI or Flask recommended) with MongoDB for storage.

☐ AI model calls must originate from the backend; the mobile client never holds the AI key.

☐ Document every build step you take so we understand your workflow.

Functional Scope

1. Single-page screening workspace that matches the demo flow : header branding, organ selector, clinical context input, JR versus SR toggle cards with usage counters.

2. Image intake and preprocessing pipeline supporting photo pick or capture plus validation before inference.

3. AI analysis request that sends model, organ, context, and image payload to your backend, which then calls Gemini or OpenAI and returns structured text with observations, preliminary diagnosis, confidence, and disclaimer.

4. Run history and session memory so every request or response is persisted with timestamps, model used, and feedback, remaining available after relaunch.

5. Usage limits per model (same daily caps as the demo) and UI indicators that reflect remaining quota.

6. Feedback capture widget per analysis that stores rating and optional notes.

7. Offline and retry UX that communicates connectivity or model errors with actionable guidance.

8. Export or share action that lets the user share the latest analysis summary.

Deliverables Checklist

☐ A Public GitHub repository with logical commit history separating client, backend, and infra work.

☐ README describing architecture diagram, setup steps, environment variables, AI model rationale, and persistence strategy.

☐ MongoDB migration or seed scripts (or clearly documented commands) for required collections and indexes.

☐ API reference such as an OpenAPI spec or Postman collection showing available endpoints.

☐ Short walkthrough video that proves the flow on simulator or device and includes a peek at MongoDB after multiple runs.

☐ Evidence of testing : at least minimal backend automated tests plus any React Native unit or component tests you consider critical.

Walkthrough Talking Points

1. Environment setup choices and overall project structure.

2. How the React Native app communicates with the Python service, including data contracts and error handling.

3. How you manage the AI provider configuration and secure the API keys.

4. MongoDB schema design for history, feedback, and quota tracking, including how persistence survives restarts.

5. Debugging and profiling techniques you used (logs, inspectors, performance checks).

6. Production-readiness touches you added such as accessibility considerations, analytics hooks, or deployment scripts.

Evaluation Signals

1. Depth of React Native knowledge, including state management, navigation, and styling discipline.

2. Backend craftsmanship in Python with clear API design, modular services, and dependency handling.

3. MongoDB schema quality for append-only history, quota tracking, and session retrieval.

4. Documentation clarity, reasoning quality, and ability to communicate trade-offs.

5. Creativity in matching the feel of the original experience without step-by-step instructions.

Submission Instructions

Email the repository link, walkthrough video demonstrating the building process and not the end result URL, and any deployment links (if hosted) with the subject line “Screening PathoAI Assignment –

”. Include total time spent and any known limitations. These materials will determine whether you move forward in the interview process.

Once done - please mail us at Extra points for any creative decisions!