Find out what you missed this month... and more:
Learning users’ habits is something that can be useful in many cases we've already covered here: smart home automation and energy consumption optimization are recent examples. Today we’re introducing a new use case: a wellness coach that is more relevant and personal thanks to craft ai. Read the article HERE
Hostabee showcased its connected device for beehives at FUTUR #IA, a CAP DIGITAL event. Hostabee, the winner of the French Tech Hackathon 2016, proposes a connected device to collect data from beehives. It uses machine learning services, especially craft ai, to enable beekeepers to remotely check the status of hives based on temperature, humidity and weather predictions. Using the Hostabee app, the beekeeper can make improved decisions, better anticipate tasks, and can also avoid wasting time in unnecessary visits as well as reduce mortality rate.
We are pleased to announce the start of project SCHOPPER1! We’ll work on craft ai core and improve its capacity to learn usage patterns from fragmentary, spatiotemporal information. Models developed during the project will be integrated in a simulation system dedicated to the validation of scientific hypothesis in the context of heterogeneous archeological data.
1SCHOPPER is a three-year R&D project funded by the ANR - Générique 2016. The consortium is composed by CERPT, CEROS, Immersion Tools and craft ai.
Welcome to Sylvain Marchienne our R&D Machine Learning Engineer
He joined us to reinforce the R&D team and prototype new Machine Learning methods to extend the craft ai API. He plans to improve the precision of the observation and detection of cycles in habits. And he already shares with us his passion for the Neufchatel French cheese 🧀
Welcome to Alexis Fontaine our Developer Advocate
He’s in charge of creating technical content for craft ai product. He will also build new demos for our clients. He never leaves his silver chain and always proudly wears it with an open collar and that's why we chose him, isn't it? So classy 😎
- Introduce the ability to use any context property as output in the agent configuration, i.e. no longer only properties with the "enum" type but now also "continuous" and time types: it is now possible to learn a tree for a numerical output (regression trees).
- Fix lower bound computation issue where split intervals were not covering the whole range of possible values.
If you are not yet registered on our API it is the right time to do so and discover the latest improvements of craft ai! Sign up HERE.
We’re still looking for new team members! Join us NOW!
- Docteur/Ingénieur(e) de Recherche en Intelligence Artificielle
- Back-End developer / DevOps
- "Developer advocate" trainee, création et communication de contenu technique Apply HERE