Far fewer conversational AI solutions this year, but far more AI automation and management solutions.
By Jérôme Malzac - Innovation Officer Micropole
The layout of this exhibition in the corridors of the Palais des Congrès is not, in my opinion, optimal for discovering all the exhibitors. The biggest players are in the entrance hall, the start-ups are relegated to a long corridor and the already well-known and well-established companies are in the opposite corridor. In short, it is always a labyrinthine game at the Palais des Congrès.
Most of the big publishers were present with attractive stands and a showcase of their flagship solutions, some interesting announcements but I won't dwell on those that can easily be found in their communication.
Last but not least, there were far fewer conversational AI solutions this year, but far more AI automation and management solutions.
Among the start-ups I selected, I will mention 4:
is a SaaS solution from Madagascar that enables the creation and deployment of artificial intelligence projects in a guided, supervised and rapid manner.
Whether the algorithms are existing on the market or specifically developed, the solution allows the creation of the entire workflow, the data feed, the actions necessary for their learning and use.
In the form of a process diagram, the construction console (via drag & drop) of the various tasks provides a clear and comprehensible graphical interface for handling and retrieval.
There are many uses in all sectors: consumption forecasts (industry), product recommendations (marketing), default predictions (banking and insurance), etc.
To be explored, to provide our algorithms with a centralised management interface which is sometimes lacking in our projects.
A no-code solution for creating conversational diagrams, which I feel is a little less powerful than our partner Kwalys.
The problem with these solutions is that they do not take into account the specificities of the conversation's display according to the channels.
The display management rules in a chatbot vs. an animated web page, the tone and latency rules of a voice bot, etc.
These solutions are all based on the same efficient and relevant technologies, but they forget the functional front end which is the purpose of their use.
No great revolution, since we are dealing here with a solution that offers analysis and dataviz models for marketing data collected online by the various tracking tools.
It is purely a customer behaviour analysis tool like others, but what attracted me was the ease of use and interfacing and the fact that the start-up is looking for partners to collaborate with to improve their models.
The promises of this platform seem very interesting to me!
Its aim: to provide large companies with a solution for the transverse management of the end-to-end data value chain!
That's all it is!
In fact, it is a platform that allows you to plan DATA projects, organise them, measure the impacts between them, quantify the risks linked to them within organisations, reduce the impact and predict the return on investment of the data and analyses of these DATA projects.
As you can see, this is a management solution that will enable companies to grasp the value of data, determine their life cycle and plan and coordinate all DATA projects.
It is both a project management solution and a business reporting and decision support solution for launching a future project.
In short, for this edition of BIG DATA & AI Paris (which by the way added AI in its name this year), no real technological revolution but a certain awareness around the (too) strong collection of non usable / non used data and the risks related to the security of these data.
A year in which data ethics, data sobriety and the principle of rationalisation began to emerge in the conferences but without any impact yet on the solutions presented. Maybe next year ... to be continued ...