Chatbots, OCR, Fullstack

WhatsApp chatbot to place and track orders for a leading lens manufacturer in Asia

The customer wanted to build a chatbot that would allow end customers, opticians, third party eyewear shops to place and track orders for lenses. Users could upload pictures of prescriptions, the chatbot would parse the data using OCR and ML to capture power, axis, spherical, and cylindrical data points. The user could then choose multiple options of lenses, enter details, and place an order. A rich web application was built to capture payments, user information, and edit orders.
2 months
Amazon AWS
AWS EC2
PostgreSQL
React
Redux
Redux Sagas
WhatsApp API
Machine Learning
Dialogflow
Firebase
OCR
Github Actions

Challenges

  • A machine learning model had to be built and trained so that it identified all the attributes from a prescription.
  • Customers could pay for an order via payment links that are generated and sent via WhatsApp.
  • Customers could interact with the product in a conversational manner to track and place orders.
  • A facade had to be build over the existing ERP system to expose a RESTful API.
System architecture diagram of a WhatsApp chatbot to place and track orders for a leading lens manufacturer in Asia

Solution

  • Wednesday setup a team comprising of a senior software engineer and a designer. The team took the idea from start to finish in 60 days.
  • The team worked with the customer to understand their ERP and build an API around it.
  • Given the short deadline the team created a strategy that allowed a fast go to market strategy by selecting tools and frameworks that allowed the customer to meet their product requirements.
Engineering, design, and growth strategies
from the digital world.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
We will not share your email with any third party. Newsletters will be delivered once a month.