VisH317 / store-assistant

Home Page:https://store-assistant-five.vercel.app

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Store Assistant

Two Main Parts: The store interface for setting up the navigation directory and the client mobile app

Mobile App

  • Detect if in a navigable store using geolocation
  • Take in voice or text input
  • Use input to return a response telling the location or availability of an item
  • Translate to languages as required
  • Future: provide availability

Mobile App Technical Specifications

  • Build using react-native
  • Need to use geolocation and a database of participating stores as well as their system prompts for setting up the store things and locations
  • Need to be able to take voice input (use whisper API probably)
  • Need to be able to process using the engineered prompt
  • Need to be translated (separate prompt)
  • Need to be converted back to voice if required (use whisper API probably)

Third part to implement: online store assistant for navigation based on preferences (chrome)

Desktop App - for store managers to set up easily Need to provide location of the store Set up prompt: mapping of aisles, relative locations of each aisle, and the items in each aisle Requires database storage, pretty much it

###Technical Specifications/Requirements

BACKEND: Database of locations => names, starting prompts (supabase) OpenAI prompt runner: includes starting prompt and loaded prompt Authentication for stores

FRONTEND (MOBILE): Home landing page Switch to chat directory when detecting a nearby location Chat to interface with the API based on provided knowledge

FRONTEND DESKTOP: Landing page => provide general information UI for store owners to create prompts => simple form with description Finished page => tell how to advertise or stuff like that

LATER CHROME EXTENSION: requires own frontend and backend

Proper Description of Stuff to do:

  • Setup supabase for the database
  • Next.JS for the desktop app:
    • Initialize the app ✅
    • Setup the prisma schema ✅
    • Setup TRPC for the REST API to interface with the prisma schema (get all stores for a user and create a new store with the data, prompt updating) ✅
    • Setup the landing page
    • Authentication ✅
    • Displaying all stores on the dashboard ✅
    • Form for signing up a store ✅
    • Updating form ✅
    • Updating styles (Proper UI design)
  • Backend Chat:
    • Setup python and deploy as supabase edge function (should be pretty simple)
  • Mobile App:
    • Setup react native
    • Setup prisma calls
    • Setup TRPC (if necessary)
    • Setup landing page
    • Setup location finder
    • Setup chat interface
  • Chrome Extension:
    • Deploy a supabase edge function
    • Plan this part out later

About

https://store-assistant-five.vercel.app


Languages

Language:TypeScript 83.6%Language:Java 6.3%Language:JavaScript 5.2%Language:Ruby 2.0%Language:Objective-C 2.0%Language:Objective-C++ 1.0%Language:CSS 0.1%