Dataiku, a global enterprise artificial intelligence (AI) company, unveiled its coding agent "CoBuild" while presenting AI internalization and governance as key factors for successful corporate AI adoption.
Dataiku held a press conference on the 24th, one day before its "Dataiku Summit" event at the Westin Parnas in Gangnam-gu, Seoul, on the 25th, to introduce its corporate AI transformation (AX) strategy and new products. Dataiku provides a platform that helps companies utilize various AI capabilities without moving cloud, server, or data. Founded in France in 2013, the company has more than 750 enterprises worldwide as clients.
Andrew Boyd, senior vice president and general manager for Asia Pacific and Japan (APJ), who attended the press conference, said, "According to McKinsey's survey last year, more than 88% of companies worldwide are using AI, but only 6% have experienced the benefits of AI adoption in terms of ROI (return on investment)," defining this situation as the "AI success gap."
Boyd added, "The most difficult and important part (for corporate AX) is AI scaling," stressing that "the value of AI shines only when AI is deeply internalized into the organization's work and actually functions." He continued, "For AI to be properly internalized, it must be easily understood by business users, controllable by risk management organizations, and stably operable by IT departments," adding that "the Dataiku platform is designed to satisfy these conditions simultaneously."
Guillaume Apper, director of product management, introduced CoBuild, which was launched on the 18th. CoBuild is a feature that converts requirements entered in natural language by field users into immediately applicable AI projects. CoBuild uses frontier AI models to identify necessary data and design workflows. It also automatically generates core components such as data pipelines, machine learning models, agents, and applications without coding.
In particular, because it visualizes all processes and outputs in graphic form, companies can carry out AI projects that fit their governance frameworks. Another feature is that it supports large language model (LLM) environments such as Snowflake, OpenAI, Anthropic, AWS Bedrock, and Google Gemini, ensuring companies' data sovereignty and model choice.
Other technologies introduced included "E2A," which converts the knowledge and judgment criteria of in-house business experts into structured reasoning logic for AI agents; "Agent Management," which integrates and manages multiple distributed AI agents on a single screen regardless of platform environment and monitors performance and risk; and "Reasoning System," which supports decision-making based on function- and industry-specific agents.
Meanwhile, Dataiku plans to introduce methods for building AI environments using its platform at the Summit event on the 25th. About 500 domestic clients, partners, and experts are expected to participate.
Kim Jong-deok, country manager of Dataiku Korea, said, "This year's Dataiku Summit Seoul will be an occasion where major domestic clients, partners, and experts participate together to see how AI is applied and scaled in actual work environments," adding, "Dataiku supports companies in stably scaling AI through an operating model consisting of people, orchestration, and governance, and we will continue to support domestic clients in developing AI into an enterprise-wide operating system rather than a one-time experiment."
