Dataiku 12 fully commits to AI-driven data processing, control and integration with OpenAI's GPT models

Dataiku 12 commits fully to AI-driven data processing, control and integration with OpenAI's GPT models

Dataiku, the platform for "Everyday AI," introduces Dataiku 12. This new release is packed with new features designed to help organizations confidently deploy AI in advanced analytics initiatives. Dataiku 12 offers deeper integration with OpenAI's GPT models and new, stronger process and governance capabilities to help organizations achieve better results with AI and mitigate risk.

Companies are striving to be at the forefront of using advanced analytics and AI, and ChatGPT and Large Language Models are on the agenda of boards of directors as a result. Companies must balance the pressure to use AI as a competitive advantage with the need to control potential risks to their reputation and operations. For many companies, the risks are overwhelming because of a lack of understanding and inconsistent processes for data, analytics and AI tools used by different teams. The new platform should address these challenges by providing trust and control over AI outputs.

New features
OpenAI GPT integration: Dataiku enables business users to integrate OpenAI's GPT models into data projects by extending datasets and executing tasks using a visual interface and clear language prompts, all while maintaining transparency and confidence in project output.

Causal machine learning (ML): Causal prediction ensures that correlation is not confused with causation when it would be detrimental to the business or other stakeholders. Dataiku democratizes these capabilities so that anyone building ML models can understand the "why" behind their results.

Universal Feature Importance: Some ML models offer limited explanations, which can make it difficult for stakeholders to rely on the results. Dataiku centralizes model explainability so that teams have a consistent way to explain models to build trust with business users.

Model revocation: A core principle of AI security is maintaining human oversight. In some cases, predictive models do not have the best or safest answer, so Dataiku allows experts, based on their actual experience, to set strict rules that keep the output of AI models in line.

Model risk project overview: With AI projects in different teams and stages of development, it is difficult to manage risk and allocate resources. With Dataiku, business and analytics specialists can easily identify and mitigate risks in AI projects, increasing confidence in project output.

Transparent automated feature development: Feature engineering can be a black box in the AI modeling process, increasing perceived risk. Dataiku gives people transparency and control over this ML engineering and helps understand where new features are coming from.