Surpassing Artificial Intelligence by facilitating machine-to-machine communication
- 10 Oct 2023
As Smile activities cover a large scope of digital services, from early conception, to design, industrialization and support, our R&D effort focuses on specific research topics. We will overview over three articles, research topics to introduce our R&D activities : Digital infrastructures resource scheduling, Machine learning and 5/6G slicing.
Machine Learning automation
Machine learning, or ML in short, is the ability for a program to learn, by analyzing data, testing an API or both, which is known as reinforcement learning. The AI (Artificial Intelligence) software domain, relies over ML techniques, powered mostly by open source solutions, such as ScikitLearn and Tensorflow.
On a day to day basis, AI supports a lot of business needs, providing recommendations, smart services or digital content to either stakeholders and end users. AI also benefits programming, with algorithms used to improve databases engines and smart tools made available to developers in modern IDEs or integration pipelines.
One of the key feature in machine learning is the ability to translate interfaces. Be them linguistic, as it is the case in the NLU/NLP domain, or visual, through the computer vision domain, machine learning provides the ability to analyze one interface or media and translate it into another. Taking the language example, that could be from a lang to another, or from an audio source to a text file. Nowadays it is easy to transform a text into an image, or even a video !
At Smile R&D we pursue an effort towards automating the ability for programs to understand one another, and possibly their ability to seamlessly interact with their users. For one program to understand another, or its human user, is to be able to translate the remote program instruction or user request into its own API.
We have been exploring the possibility to automate knowledge discovery and APIs translation through few approaches : media to media transformation, 3D visualization, smart services chaining ; and for various use cases : e-commerce products catalog enhancement, helpdesk issues qualification, distributed systems supervision…
You can find most of our open source research results within the SMILE github group : https://github.com/Smile-SA.
If you want more information, feel free to drop us a mail at email@example.com.