Skip to content

FabricStore.md

Latest commit

 

History

History
71 lines (48 loc) · 3.28 KB

FabricStore.md

File metadata and controls

71 lines (48 loc) · 3.28 KB
Raw

Internship Assignment: AI-Powered Fabric Sourcing Assistant

Objective: Build a web application featuring a "Voice-to-Form" interface designed for fabric merchants. The form will allow users to describe fabric details (type, GSM, color, etc.) via audio. The system must transcribe the audio, extract key entities using a Python backend, and dynamically populate the form for final manual review and submission.

Requirements:

1. HTML Structure & UI

  • Create a clean, responsive form with the following 5 fields:
    • Fabric Type (e.g., Cotton, Silk, Polyester)
    • GSM (Grams per Square Meter)
    • Color/Pattern
    • Quantity Required (in meters/yards)
    • Delivery Timeline
  • Include a prominent "Speak" button with a recording indicator.
  • The UI must be professional and minimalist (OpenGraph compatibility is required for the final view mode).

2. Voice Recording

  • Implement audio recording functionality (Web Speech API or MediaRecorder API).
  • Enforce a strict 1-minute limit for recordings.
  • Provide visual feedback (e.g., a pulsing icon or timer) while recording is active.

3. Dynamic Form Filling (The AI Core)

  • Backend: Use a Python backend (FastAPI or Flask managed via Poetry).
  • Processing: Transcribe the audio and use a lightweight LLM or NLP logic to extract the 5 fabric entities.
  • Logic: Auto-fill the relevant fields. If a specific detail (like GSM) isn't mentioned, that field must remain empty for the user to fill manually.

4. Edit & Validation

  • All auto-filled fields must remain fully editable by the user.
  • Implement client-side validation to ensure numeric fields (GSM/Quantity) receive correct data types before submission.

5. Submit and View

  • Store the submitted data (JSON or SQLite).
  • Upon successful submission, redirect to a "Summary View" page that displays the fabric request details in a clear, read-only format.

6. User Experience & Error Handling

  • Handle scenarios where the microphone is blocked or the transcription service fails.
  • Ensure the transition from "Audio Processing" to "Form Populated" is seamless and fast.

7. Optional Enhancements (Bonus)

  • Audio Playback: Allow the user to replay their recording before hitting submit.
  • Advanced Extraction: Use Pydantic models for structured data extraction on the backend.
  • Styling: Use a modern CSS framework (Tailwind or similar) to ensure a high-quality feel.

Submission Guidelines:

  1. Codebase: Push the complete code to your GitHub repository.
  2. Documentation:
    • A README.md with clear setup instructions (include poetry install steps if using Python).
    • A brief technical document explaining your choice of transcription library and how you handled the entity extraction.
  3. Timeline: 1 Week from the date of receipt.

Evaluation Criteria:

  • Accuracy: How well the AI identifies fabric-specific terms and maps them to fields.
  • Code Quality: Clean, scalable, and modular code with minimal but effective comments.
  • Production Readiness: Proper environment management and error handling.
  • UI/UX: Intuitive design and smooth interaction flow.

© 2026 Recursive Zero Pvt Ltd, All rights reserved.