Overview

The client is the USA’s best-known automobile selling company, mainly focusing on selling second-handcars online.The client provides a fully automated service scheme for online car buying procedure. Over the last one and half decade, the client has helped millions of owners sell their cars online. Even today, the client’s brand is the first and foremost choice for USA’s customers to sell their cars online. The client’s business model has been acclaimed as a pioneer one; thereforethe clientwants successfully implement the latest technology that won’t interrupt the business and will give customers a hassle-free buying experience.

Domain
Machine Learning

The Requirement

  • The Client’s website serves as a platform to showcase used cars that are ready for sale.
  • The client’s website leveragesthree times the reach with customers than the opponents’ website do.
  • Require a convention rotating 360° photo studio that automatically captures and process 16 standard pictures of every car in their stock.
  • The requirement was to capture high-resolution photographs, remove the shadows, and provide additional effects with comparable colors to meet standard image specification.
  • The client also wanted to leverage time differences with an expert photo editor, so the images received at the end of the day is ready by or before the next day.
  • The client also wanted to analyze a dataset of images, covering various cars with the right budget model specifications.

The Challenges

  • Urgent need of automobile photo editing requirement for thousands of images due to the nature of client’s business required us to work at quick short turn-around times
  • The client wanted the images to be edited 24*7 similar to outsourcing companies do.
  • The assigned team strength was not adequate at the point of time to fulfill the client’s requirement.

The Solution



  • As we follow Agile and Scrum methodology, so we decided on a phased approach wherein our skilled experts worked on a delivery schedule by gradually increasing turnaround volume
  • Our team worked on a solution where the dataset contains a huge number of car pictures as .jpg documents by editing 1500 images fortnight to meet the client’s expectation within the requested deadline.
  • Every individual car was precisely captured with 16 pictures, and each of them was taken from different angles. Every carwas given a unique id and photos were named by cid_01.jpg, cid_02.jpg, cid_16.jpg and the rest were in a similar format.
  • Our team members take the initiative and worked overtime to meet the project deadline. We also hired and trained 07 new professional image editors, to fulfill the criteria of 24/7 service

The Development


Training Phase

  • We’ve used Dense layer, Convolutional Layers, IO Layer,Training phase was performed on AWS Sagemaker service.

Model Deployment

  • AWS Sagemaker hosting service

Achievement

  • By implementing this solution on cloud-based infrastructure. End-user get a quick response and reduce the cost for handling support.
  • The useris able to get clear thoughts and visibility for their target product

Technology Stacks

  • Cloud Infrastructure
    AWS Sagemaker
    AWS ECS(Application Server)
  • Technologies
    Python
    Keras
    Nodejs
    React
    Postgres SQL
Team
acr

1

React Developer

acr

1

Node Developer

acr

1

HTML Designer   Graphic Designer

acr

1

Deep Learning Developer

acr

1

Cloud Developer

acr

1

Quality Analyst

The Result

  • We kept our promise by delivering 100% accurate and professional quality images for superior client satisfaction
  • The client was indeed satisfied with our strict turnaround times as we always adhere to the deadline
  • Delivering the project in the mentioned deadline, the client willingly signed further ongoing contracts with us
  • The team is delighted to be assisting the client in carrying out the desired result in car sale in the USA, as we take equal pride in helping the client with the desired outcome.
Contact us