Building, training and deploying a ML model to classify objects based on photos

Overview

Industry

Location

Manufacturing

Poland

Technology Used

.Net

Python

Client background

A branch of an international group that recycles steel scrap and non-ferrous metals, this client plays a significant part in the green transformation. The group provides high purity recycled materials that can instantly decarbonize products.

Project overview

Our team was responsible for building, training and deploying a machine learning model based on photo data sets provided by the client. The model was trained and hosted in an AWS cloud. Additionally, our team built a mobile application that integrated with the model, enabling users to take scrap metal photos and classify images in real time. The final part of the project was a web application for managing classifications and the entire junkyard.

Results

Efficient production environment
Composing a complete production environment with a role-based administration panel
Enhanced image classification
Implementing a ML model for actual image classification
Elevated communication
Using Rest API for communication between components
Better image integration
Providing integration with the ERP system

We'd love to hear from you!

Fill out the form - we'll get back to you as soon as possible

More case studies

Software and IT services

Executing a cyber security audit of a LegalTech platform for integrity and compliance

Software and IT services

Reducing hardware footprint through router-to-firewall migration

Financial services

Increasing financial services flexibility with a comprehensive cloud migration

Biotech and Healthcare

Crafting a secure electronic and postal voting app for doctors

Telecom

Deploying a modern NMS platform for alarm management across a telco’s network

Newsletter