Treasure AI Agentic Experience Platform & CDP

Treasure AI, formerly Treasure Data, is the business name of Treasure Data, Inc., a Delaware corporation with its dateline in Mountain View, California. Co-founded by Kazuki Ohta and Hironobu Yoshikawa, it has been building an enterprise customer data platform since 2011 and has offices in New York, Tokyo, London, Nice and Vancouver. It helps large consumer brands unify customer data from many sources and act on it across marketing, advertising and service.

The platform has three layers. The data layer is a complete or composable CDP with identity resolution and integrations with Snowflake, Databricks and BigQuery. The AI layer includes Treasure AI Studio, a conversational workspace with more than 50 skills that builds segments and journeys from unified profiles, plus an AI Agent Hub and the Treasure Code command-line tool. AI suites for engagement, personalisation, creative, paid media and service activate the results. Users include Sony, Yamaha, AB InBev, Canon, Nestlé, Subaru and Shiseido.

• Enterprise CDP: Offers complete or composable customer data platform deployment.
• Identity resolution: Merges records from many sources into unified customer profiles.
• Treasure AI Studio: Builds segments, journeys and campaigns through conversation with more than 50 skills.
• AI suites: Activates engagement, personalisation, creative, paid media and service use cases.
• Treasure Code: Lets data teams manage the CDP from the command line and coding assistants.
• Governed agents: Requires human approval and includes evaluation and hallucination controls.

A multinational retail or consumer goods group in Vietnam can unify loyalty, e-commerce and store data into one customer view. A cosmetics brand can let marketers create segments and campaigns by conversation. An automotive or electronics company can link CDP audiences to paid media to reduce advertising waste.

Marketing and data teams can turn unified customer data into segments and campaigns through a conversation, with human approval before activation. The company states up to 50% lower costs, a 2.5 times conversion lift and up to 90% less CDP operating effort with Treasure Code. Case examples include 14.5 times ad efficiency for Subaru, USD 1 million saved for Six Flags and more than 2,000 data sources unified for AB InBev.

Large brands are moving from CDPs that only store data to platforms where AI agents build audiences and campaigns directly from unified profiles. Composable CDPs that sit on top of cloud data warehouses are gaining ground. In Asia, where Japanese and multinational retail groups run loyalty programmes across many markets and channels such as LINE, demand is rising for governed AI on top of first-party data.

Treasure Data, Inc. has built its CDP since 2011 and rebranded as Treasure AI in April 2026. It reports more than 400 enterprise customers including 80 Forbes Global 2000 companies, over 2 billion profiles managed, presence in more than 75 countries and over 550 employees. SoftBank representatives sit on its board. It was named a Leader in the IDC MarketScape for AI-enabled B2C CDPs in 2026.