Slide Background

The world's first sound-based AI Multi-Agent system for Advanced Quality Control and Predictive Maintenance


⏱️ Avg. Reading Time: 5 min

Slide Background

Hi there !

Looking to improve your production efficiency with AI? 

Leading manufacturing companies struggle with costly Big Data services, at the risk of compromising the control, profitability and scalability of their initial investment. 

As for medium-sized companies ? They're watching the train pass ! ... because these investments are far too costly

We have the solution! 

Or rather, we've turned this world of AI experts into a product: iDetect-4.0, whose efficiency and simplicity are attracting more and more fans. 

Join them! Here's a sneak preview. 

Alain Fanet

Edge AI or not ? 

In the industry, the execution of machine learning  & Deep learning models at the Edge comes with several benefits over conventional cloud-based AI:

  • Low Latency
  • Reduced Power Consumption
  • Improved Environmental Performance and safety
  • Network Bandwidth Efficiency
  • Strong Privacy

But the built-in models can only make inferences and the reliance on cloud/edge training poses new challenges

  • Challenges
  • Solution
  • How it works
  • Benefits

Industry Challenges

To deploy AI Deep Learning  in  Production

30% of AI projects stall during training data preparation

... due to large datasets required

Only 1% of the real data is part of the training dataset

... Edge AI models experience performance drift due to unexpected field changes.

Close to 60% of costs stem from models maintenance

... As result, less than 10% of industrial AI projects progress beyond the Proof of Concept (PoC)

1000 x

faster than competive solutions

Our Edge AI Learns  ...

An innovative algorithm, learning "in-context", our AI solution dynamically adapts to changes and variations locally (on the device) in real time (seconds), adapting to model drift, generalizing to different data types and learning efficiently from limited live data captured, without relying on a large-scale data infrastructure at the edge.


Step 1: Small Data Collection

The system collects automatically few (tens) live data from machines and sensors.

Step 2: Automatically build the AI deep learning model

Our cutting-edge AI algorithm, automatically, creates Deep Learning model and turn it in production in second.

Step 3: Continuous Model Performance improvement

DavinSy works with professionals or machines in real-time to adapt to unexpected changes, rebuilding models directly from the device


Reduced project launch risk

3x reduction in time-to-market and market costs

Greater Economic Efficiency 

7x Net Profit Increase

Increasing the value of data of interest directly in the field

90% of raw data volume can be reduced by DavinSy at the Edge

Valuing human expertise 

Maximize model performance by drawing on the best professional know-how.

Slide Background

What we do

Thanks to our groundbreaking deep learning engine, we met these challenges with a simple and effective solution: iDetect-4.0. This is the world's first AI sound-based solution that provides manufacturers with a comprehensive approach compatible with multiple manufacturing sectors.

Our innovative AI multi-Agent solution enables early detection of faults or degradations that may not be visible to the naked eye. By identifying sound issues proactively, iDetect-4.0 allows for timely interventions, continuously enhancing production efficiency while minimizing downtime and costly repairs and defects.

Our Product Offering


AI-based Quality control, predictive maintenance, read more... 

iDetect-4.0 offers advanced quality control and predictive maintenance capabilities through AI Multi-Agent sound based system. This AI operating system is designed to run seamlessly on Edge industrial PCs, offering advanced capabilities in maintaining deep learning AI models, managing data of interest and supporting quality processes. Each operating mode (quality, maintenance) is powered by a dual deep learning AI Agent (DavinSy). One creates deep learning models used in production, and the other continuously improves model performance based on corrections made by the operator and validation by the quality engineer. Ultimately, iDetect-4.0 empowers autonomous Machine Learning Operations (MLOPs), eliminating the need for additional infrastructure while significantly enhancing manufacturing processes.

DavinSy AI Operating Agents

Orchestrate entire AI workflow and ressources,

Those Agents are operating as a suite of micro-services,  up to five DavinSy Agents to orchestrate the entire AI workflow, tailored for classification and pre-processing for extraction features using Sound Analysis. These agents function autonomously and can dynamically create AI models from live data, in seconds. 

Maestro Platform

Streamlines design, deployment, and monitoring,

A configuration platform designed to streamline the design, deployment, and monitoring of IDetect product deployed. The configuration is done through a user-friendly, no-code approach. All features ensure the adaptability of iDetect-4.0 across diverse machines and environments, enabling swift model porting and adaptation within a matter of days. Maestro lets you create collaborative workspaces and shared projects between internal departments, with different rights for different members, but also including third-party external professional services.

Unique Selling Points

Adaptability and Scalability

Adaptable to Diverse Machinery and Environments: Our multi-AI-agent sound processing system adjusts models rapidly, ensuring continuous adaptation without machine downtime


Fast Track to Operational Performance

Rapid Onboarding: Starts with Few Sound Records or No Data and achieves operational performance within days, boasting 98% accuracy, 0% false negatives, and less than 2% false positives.


Cost-Efficient Edge Architecture

Cloud-Free Architecture: Enhances Data Confidentiality and Security.

Qualitative Dataset Generation: Optimizes Storage and Reduces Raw Data Volume.


Productive & Collaborative Digital Workspaces

Collaborative Software-as-a-Service: Seamless Integration with Existing IT Infrastructure.

No-Code Configuration: Easily configure data acquisition, signal processing, modes, user access, and operator interface, facilitating efficient Design & Performance Tuning Phases.


Business Model and Client Benefits

Maestro is a subscription-based SaaS, and deployment of iDetect 4.0 is license-based.

Significant Reductions in Deployment Time and Costs: Achieve a 3x reduction in time and costs-to-deploy.

Enhanced Data Management: Benefit from a 90% reduction in raw data volume and keep data of interest.

Slide Background


160 Avenue de Fes, 
34080 Montpellier

+33 6 63 44 61 33



Would you like to know more?