Dataiku 11: Top features that will accelerate enterprise AI projects

We’re excited to bring Transform 2022 back to life on July 19th and virtually July 20-28. Join AI and data leaders for sensible conversations and exciting networking opportunities. Register today!


New York-based Dataiku, which provides centralized solutions for the design, deployment and management of enterprise artificial intelligence (AI) applications, has released version 11 of its Unified Data and AI platform. The update, which will normally be available in July, focuses on the promise of “everyday AI” and offers new capabilities to help data experts handle more elaborate AI projects, but non-technical business users can easily connect with AI. Enables it. Improved workflow among other benefits.

“Expert data scientists, data engineers and M.L. [machine learning] Engineers are some of the most valuable and sought after jobs today. Yet often, talented data scientists spend most of their time on low-value logistics such as setting up and maintaining the environment, preparing data, and putting projects into production. With the extensive automation built into Dataiku 11, we’re helping companies overcome the frustrating engagement so that companies can quickly make more use of their AI investment and ultimately create a culture of AI to transform the industry, “said Clement Stenac, CTO and co-founder of Dataiku. Said.

Below is a compilation of key capabilities.

Code studio with experiment tracking

Code Studio in Dataiku 11 provides AI developers with a fully managed, separate coding environment in their Dataiku project, where they can work using the IDE or web application stack of their choice. This solution gives AI developers a way to code how comfortable they are while adhering to their company’s policies for analytics centralization and governance (if any). Previously, anything like this had to go for a custom setup with increasing costs and complexity.

The solution also comes with an experiment-tracking feature, providing developers with a central interface for storing and comparing all Bespoke model runs run programmatically created using the MLFlow framework.

Seamless Computer Vision Development

To simplify the resource-intensive task of developing computer vision models, Dataiku 11 brings built-in data labeling framework and visual ML interface.

The former, as the company explains, automatically criticizes large amounts of data – a task often handled by third-party platforms such as Tasq.ai. The latter, meanwhile, provides an end-to-end, visual path for general computer vision tasks, enabling both advanced and novice data scientists to tackle cases involving the use of complex object discovery and image classification, from data preparation to model development and deployment. Until.

Time-range forecast

For business users, especially those with limited technical skills, it can often be difficult to analyze historical data and create robust business forecasting models for decision making. To address this, Dataiku offers 11 built-in tools that provide a no-code visual interface and help teams analyze temporal data and develop, evaluate and deploy time-series forecasting models.

Feature Store

The latest release also brings a feature store with new object-sharing flow to improve organization-wide collaboration and accelerate the whole process of model development. According to the company, the capacity data will give teams a dedicated zone to access or share reference datasets with curated AI features. This will prevent developers from re-engineering similar features or using redundant data assets for ML projects and prevent inefficiencies and inconsistencies.

Outcome optimization

Teams often use manual trial and error (if any) methods to provide business stakeholders with actionable insights that can help them achieve the best possible results.

With the result optimization coming as part of Dataiku 11, the whole process will be automated. In essence, it will automatically take into account user-defined constraints and find the best set of input values ​​that will yield the desired results. For example, it may determine what changes the manufacturer may make in the factory conditions to achieve maximum production yields or what adjustments in the bank customer’s financial profile are the least likely to default on a loan.

Other capabilities

Among other things, the company has introduced tools to improve monitoring and control over model development and deployment. This includes an automated tool for generating flow documents and a central registry that captures snapshots of all data pipelines and project artifacts – for pre-production review and sign-off. The company will also provide model stress tests, which will examine model behavior in real-world deployment situations before actual deployment.

Similar Posts

Leave a Reply

Your email address will not be published.