THINGS THAT THINK.
Stop model drift with on device continuous deep learning
Davinsy, the world’s first cloudless artificial intelligence of Things (AIoT) system.
Data driven AIoT control system
Voice-Sound AIoT control system
Development environment for DavinSy based AIoT system
Getting AI on board has never been easier.
ZERO DATASET. More than 30% of industrial AI projects are stuck in collecting and building a data set that should reflect the accuracy of the final application in its real environment. We solved the problem of starting to learn with little or even no data set.
We learn from real data captured live by the embedded system in a fraction of a second without the need for giant infrastructure such as cloud-based computing.
We build the dataset while the embedded system is running in the field, giving it a high level of adaptability to context changes.
The company has successfully broken through the “sound barrier” of AI workflow complexity and is now launching the world’s first data-driven virtual pattern control system, called DavinSy, to market. It is now possible to continuously learn from “live” data to guarantee the performances of embedded AI models just-in-time.
Using your existing endpoint HW (no special HW or accelerator needed) and not transferring any data to the cloud, DavinSy minimizes adoption and operation costs. DavinSy has nearly zero carbon footprint.
From startus-insights : Bondzai amongst 10 hotest AI startups addressing industry 4.0.
Pitching at BPI to digital experts and European investors, Bondzai wins the digital award in France Transfer Invest by Tech Tour. ...
Bondzai took advantage of the Global Industrie conference to meet its customers. It presented its product DavinSy which capabilities to learn very...
BLOG_ latest posts
Davinsy, our autonomous machine learning system, integrates at its core Deeplomath, a deep learning kernel continuously learning from real-time live incoming data.
Avoid catastrophic forgetting with Bondzai’s event-driven continuous and real-time embedded few-shot learning
This paper describes how Bondzai’s event-driven continuous learning permits to avoid catastrophic forgetting problems in deep neural networks.
This paper describes how we proceed with continuous machine learning (ML) at Bondzai and why we think this is a ‘must-have’ when targeting industrial applications…