Skip to content
logo-senswork-2023
Neuralyze Wortmarke
  • Platform
    • Neuralyze Desk
    • Neuralyze Core
    • Neuralyze Edge
    • Licensing
  • Use Cases
    • Surface Defects
    • Completeness Check
    • Dimensional Stability Check
    • Text/Label Verification
    • Anomaly Detection
    • Feasibility Study
  • Technology
    • Deep Learning & CNNs
    • Data Sovereignty & On-Premise
    • No-Code Workflow
    • MLOps & Monitoring
    • Industrial Research as a Foundation
  • Solutions
    • For Production Managers
    • For System Integrators
    • For Quality Engineers
    • For Education and Research
  • Start
    • Request Neuralyze
    • Contact Sales
  • English
  • Deutsch
En
PLATFORM

Neuralyze® as Platform

Neuralyze® covers the entire machine learning workflow: from data collection through training to production. You train neural networks using your own product data without any coding knowledge and without a data science department.

The platform consists of three components:

Neuralyze® Desk for training and project management, Neuralyze® Core as the inference engine, and Neuralyze® Edge for deployment directly on the production line.

Neuralyze_Keyvisual_RGB_Website_1c_lila
Neuralyze_Desk

Neuralyze® Desk

Training and Project Management

  • Learn more
Neuralyze_Core

Neuralyze® Core

The Inference Engine

  • Learn more
Neuralyze_Edge

Neuralyze® Edge

Inference Directly on the Machine

  • Learn more

NEURALYZE® DESK

Training and Project Management

Neuralyze® Desk covers the entire ML workflow in a single desktop application

Collecting and Evaluating Data 
The first step is to define the testing task: Which features should be detected? Image data from production is systematically collected and checked for suitability.

Sorting and Cleaning Data
Not every image is suitable for training. Relevant images are selected, and quality issues are identified and removed.

Annotation (Labeling)
The relevant features in the image data are marked and labeled. The quality of the annotation determines the quality of the model.

Data Augmentation
Synthetic expansion of the dataset by varying position, rotation, and image parameters. This increases robustness against real-world variations.

Train the model
You choose the network architecture; Neuralyze optimizes the parameters. The results of each variant are automatically compared.

Validate the model
Test the model on data it has never seen before. This will show whether it can generalize or whether it has merely memorized the data.


NEURALYZE® CORE

The Inference Engine

Neuralyze® Core executes trained models. A decision is made based on an image

Inference

Inference means that the trained model evaluates an image and delivers a result. The speed of this decision determines whether the system can operate at the production cycle rate.

Types of Assignments

  • Classification – The image is assigned to a category: Good, Bad, Type A/B/C. Fast and reliable for binary decisions.
  • Object Recognition –  Multiple objects in the image are located and classified. Ideal for counting or when multiple defects occur per image.
  • Segmentation –  Pixel-precise classification. The result is a precisely defined mask, which is relevant for identifying related areas and for measuring.
  • Hybrid Strategie – Enables the implementation of complex inspection scenarios.

NEURALYZE® EDGE

Compact Integration into the System

Sensor, data analysis, and decision-making all in one device on the production line

Neuralyze Edge brings inference to the machine directly on the production line, without network latency, without the cloud, and on edge devices that serve as both sensors and analysis hardware.

Training and inference are clearly separated: Training takes place under controlled laboratory conditions on high-performance hardware (Neuralyze® Desk), and the optimized models are then transferred to GPU-based edge devices.

Deployment Scenarios

  • Embedded systems directly in the test system
  • Edge servers in the production hall
  • GPU workstations for multiple parallel camera lines
Fast execution is critical to the cycle time of the testing process. Edge deployment minimizes latency and enables real-time decisions in sync with the production line.
LICENSING

Licenses & Pricing

Three licenses for different levels of access.  Choose the license that best fits your needs.

License Comparision

  Explore Trial Pro
Price Free for the education sector Free 7.000 €
Duration 2 Years 30 Days Unlimited
Target Audience Education, Research Evaluation Production
Features Only Neuralyze Desk Complete Complete
Commercial Use No No Yes
License Type Host-based Host-based Host-based or USB

 

Lizenzvergleich

Explore Trial Pro

Preis
Kostenlos

Preis
Kostenlos
Preis
7.000 €
Laufzeit
Unbegrenzt
Laufzeit
30 Tage
Laufzeit
Unbegrenzt
Zielgruppe
Bildung, Forschung
Zielgruppe
Evaluation
Zielgruppe
Produktion
Funktionsumfang
Voll
Funktionsumfang
Voll
Funktionsumfang
Voll
Kommerzielle Nutzung
Nein
Kommerzielle Nutzung
Nein
Kommerzielle Nutzung
Ja
Lizenzform
WIBU-Dongle
Lizenzform
WIBU-Dongle
Lizenzform
WIBU-Dongle (File/USB)

 


Explore
Free for Education and Research
 

Full functionality of Neuralyze Desk, limited to 2 years, free of charge. For research institutes, universities, and trade schools. Already used in the curriculum of Bavarian AI technical schools.


Trial
Try it for 30 days with no obligation
 

30 days of full access for evaluation using your own data. No automatic renewal, no hidden costs.


Pro
For Production Use


Price: 7.000 € (one-time)

Unlimited license for commercial use. Technical support. One additional Explore license included. Offline-capable.

Contact

senswork GmbH
Innovation Lab

Friedenstraße 18
81671 Munich

neuralyze@senswork.com

We share insights with our audience across social media. Check out our latest posts:

  • Datenschutz
  • AGB
  • Disclaimer
  • Impressum
  • Cookie Einstellungen

 

© 2026. All rights reserved