Ai+ Training

Implementing artificial intelligence, Explainable AI(XAI), and specific Transformer NLP models in a real-life project can prove to be quite a challenge for an unprepared team. When Python, Pytorch, Tensorflow, Trax, and other AI tools meet reality,  special knowledge and skills are required.

This presentation will take you from theory to practice. It will show you how to implement AI when the mathematics of AI conflicts with the real world's physics: lack of data, lack of computer power, human expertise, and funding. 

Learning artificial intelligence is tricky. Understanding that Transformers have made Recurrent Neural Networks(RNNs) and even LSTMs obsolete is challenging. Once somebody has made it through these obstacles, a sense of satisfaction emerges.

However, on day one of a real-life corporate project, more often than not, an AI Specialist must face critical problems: obtaining data, enforcing privacy regulations, adapting an AI model to the reality of a customized solution, and AI-related obstacles. 

Then comes the tsunami of classical processes and algorithms that constitute the mandatory components of a real-life pipeline: preparing the data, choosing the right machine, finding the right model, designing the user interface, and more, much more!

The presentation will illustrate real-time and delivery for millions of online consumers and thousands of targeted videos on streaming platforms such as Netflix, Amazon Prime, Disney+, and YouTube. This automated e-business system will collect consumer features in real-time and make highly efficient suggestions.

The presentation will explore an AI pipeline from A to Z with notebooks in Python, Pytorch, and the main frameworks.

The pipeline will begin by collecting consumer feature data with machine learning and deep learning algorithms such as a Restricted Boltzmann Machine(RBM). Explainable AI(XAI) will ensure that the dataset is ethical and legal using WIT, SHAP, and other XAI algorithms. Finally, the presentation will introduce the mind-blowing NLP Transformer models to predict consumer behavior and create automatic near-human real-time dialogs.

At the end of the presentation, you will fully understand the factors involved in a real-life implementation of artificial intelligence algorithms in a corporate pipeline with an unlimited number of users.

Local ODSC chapter in Paris, France