// AI Recognition Startup from Austria

AI-powered brick recognition for faster inventory workflows.

Software-Hub develops an AI-based Brick Recognition & Counting System that detects, counts and classifies mixed plastic building bricks from camera images. The goal is to help collectors, resellers and small businesses automate sorting and inventory workflows.

View product View MVP mockup Contact us

// Product

Brick Recognition & Counting System

The product is designed to turn camera images of mixed building bricks into structured inventory data. Instead of manually counting parts, users can upload or capture images and receive detected objects, estimated quantities and classification results.

Current MVP focus

We are building the first version of the recognition pipeline: image capture, preprocessing, object detection, duplicate-safe counting, part classification, result correction and exportable inventory data. The first target is a practical MVP for real-world brick sorting workflows.

Capture

Images are captured from a fixed camera setup or uploaded through a web interface.

Detect

The system detects visible brick parts and separates them from the background.

Count

Detected parts are counted while avoiding duplicate counting across processing steps.

Export

Results are converted into structured inventory data for dashboards, APIs or CSV exports.

// MVP Mockup

How the product workflow will look

The first MVP is planned as a web-based workflow: users upload or capture an image, the recognition pipeline detects visible parts, and the dashboard returns counts and structured inventory data for review.

Input image

Camera image / upload area
white background + mixed brick parts

Recognition result

Detected parts128

Duplicate-safe countenabled

Classificationshape + color

ExportCSV / API

Prototype test showing detected brick parts with bounding boxes and estimated counts

Prototype detection output: early test images with bounding boxes around visible brick parts and estimated object counts per image. This supports the MVP recognition pipeline and will be improved with better datasets, classification and duplicate-safe counting.

// Technology

Technology areas

Software-Hub combines software engineering, computer vision and cloud infrastructure to build a scalable recognition workflow.

👁️

Computer Vision

Detection and classification of plastic building bricks from images using image preprocessing and ML-based recognition.

Object DetectionImage ProcessingClassification

🧠

Machine Learning

Model experiments, dataset preparation, evaluation and iterative training for better recognition quality.

MLTrainingEvaluation

☁️

Cloud Architecture

Scalable APIs, storage and asynchronous processing pipelines for image uploads and recognition jobs.

APIsQueuesStorage

📊

Inventory Dashboard

A web interface for reviewing detections, correcting results and exporting structured inventory data.

DashboardCSV ExportReview UI

// AWS Use Case

Why we plan to use AWS

AWS credits would help us build and test the MVP infrastructure without overcommitting early-stage startup budget. The architecture is planned around scalable image processing, serverless APIs, data storage and monitoring.

🗂️

Image & dataset storage

Store uploaded images, generated annotations, model outputs and inventory exports in a structured way.

Amazon S3Lifecycle Rules

⚡

Serverless backend

Run lightweight API and job-triggering logic for uploads, result retrieval and user workflows.

AWS LambdaAPI Gateway

🧪

Model experiments

Test AI and ML workflows for recognition, classification and future model improvements.

Amazon BedrockML Workflows

🛢️

Application data

Store users, recognition jobs, correction history and inventory result metadata.

Amazon RDSPostgreSQL

📦

Containerized processing

Package heavier recognition jobs as containers for repeatable processing and future scaling.

ContainersECRECS

📈

Monitoring & cost control

Track system health, application logs and AWS spending while the MVP is developed.

CloudWatchAWS Budgets

// Customers

Target customers

The first target market is people and small businesses that handle many mixed building bricks and need faster counting, sorting and inventory preparation.

🧱

Collectors

Private collectors who want to organize, identify and count mixed parts faster.

🏪

Resellers

Small resellers who need structured inventory data for used brick lots.

📦

Sorting workflows

Workflows where parts are poured onto a fixed background and analyzed by camera.

// Product status

MVP development roadmap

Software-Hub is currently in the early MVP phase. The goal is to validate the recognition workflow first, then improve accuracy, usability and deployment scalability.

Phase 1 — Recognition pipeline

Build the first image preprocessing, object detection and duplicate-safe counting workflow for controlled camera images.

Phase 2 — Review dashboard

Create a web interface where users can review detections, correct results and export inventory data.

Phase 3 — Cloud MVP

Deploy scalable storage, APIs, background processing and monitoring for real test users.

// Company

About Software-Hub

Software-Hub is an early-stage software project based in Styria, Austria. We develop custom software solutions and are currently focused on an AI-powered Brick Recognition & Counting System.

The project is in MVP development. Our technical background includes full-stack web development, backend systems, cloud infrastructure, container workflows and performance-focused software engineering.

For inquiries, partnership discussions or technical questions, contact us at office@software-hub.org.

Austria

Company location

MVP

Product development stage

Recognition technology focus

B2B

Reseller and sorting workflows

// Imprint

Legal notice

This imprint provides basic contact and operator information for Software-Hub.

🏢

Operator

Software-Hub
Styria, Austria

📍

Address

Farch 2a
8741 Weißkirchen
Austria

✉️

Contact

Email: office@software-hub.org
Website: https://software-hub.org

⚖️

Responsibility for content

Responsible for the content of this website: Software-Hub. The information on this website describes an early-stage MVP project and may be updated as the product develops.

// Legal & Contact