Teradata unveils AgentBuilder to speed enterprise AI adoption

Teradata has introduced AgentBuilder, a suite of features aimed at accelerating the development and deployment of autonomous, contextually intelligent AI agents for enterprises.

AgentBuilder launch

AgentBuilder is designed to enable enterprises to quickly build and manage autonomous AI agents that combine contextual knowledge, domain expertise, and hybrid deployment flexibility. According to Teradata, the new suite leverages open-source frameworks and is powered by the Teradata AI and knowledge platform, allowing teams to design, operationalise, and oversee multi-agent systems drawing upon Teradata Vantage's data, analytics, and infrastructure.

The platform also includes pre-built Teradata Agents, which are task-driven templates developed to streamline implementation and deliver outcomes for complex, domain-specific scenarios.

Addressing operational challenges

The adoption of agentic AI within enterprises has been hindered by issues such as fragmented data that can impair reliability, lack of embedded business knowledge, performance and cost challenges linked to prompt-driven workloads, and concerns with governance that can complicate deployment.

Teradata said its AgentBuilder addresses these hurdles by integrating contextual knowledge, domain expertise, and hybrid infrastructure, enabling organisations to operate multiple secure, autonomous AI agents across both cloud and on-premises environments. The suite's agent templates are engineered to include business-specific logic, embedding relevant context directly into agent workflows to aim for more reliable outputs. These solutions are based on Teradata's platform, AI, and analytical capabilities, including the recently unveiled Model Context Protocol (MCP) Server.

Sumeet Arora, Chief Product Officer at Teradata, said, "AgentBuilder represents meaningful progress in advancing agentic AI for the autonomous enterprise. By combining the flexibility of open-source frameworks with Teradata's AI and knowledge platform and our MCP Server, which provides deep semantic access to enterprise data, we're helping organisations build intelligent agents that are not only autonomous and scalable, but also deeply aligned with their business goals, governance standards, and domain expertise."

Arora continued, "Add to that our seamless support across cloud and on-premises environments, and we're delivering a level of flexibility, integration, and contextual intelligence that sets Teradata apart. This is not just about data - it's about delivering trusted, transparent, and complete knowledge to power the next generation of AI."

MCP Server and technical foundations

The MCP Server, which forms the backbone of AgentBuilder's functionality, supplies developers and AI practitioners with a collection of prompts, components, and resources to facilitate access to the Teradata Vantage platform. With MCP, agents can query, reason, and act with higher accuracy. The integration of MCP into AgentBuilder aims to ensure that agents remain context-aware, secure, and compliant with enterprise standards. Teradata states that this integration is intended to boost development speed, dependability, and help businesses realise the benefits of agentic AI at scale.

Open-source interoperability

AgentBuilder's early release will be compatible with open-source frameworks such as Flowise and CrewAI, with support for LangChain and LangGraph expected in future updates. These frameworks provide building blocks needed for constructing the workflows, memory, reasoning, and coordination essential to autonomous systems. When combined with Teradata's scalability, governance, and performance attributes, these tools are intended to expedite the introduction of AI agent applications that can learn and adapt within business contexts.

Pre-built agent examples

Teradata Agents, included with AgentBuilder, are pre-configured templates aimed at guiding teams in adopting agentic AI for specific tasks. For instance:

• The Teradata SQL agent translates natural language queries into SQL statements, can identify schema and table structures, and refines or optimises query statements for use on Teradata data warehouse resources. It is designed to function with open-source frameworks within multi-agent systems.
• The Teradata data science agent constructs executable machine learning pipelines from natural language instructions, using large language models and Teradata MCP tools to manage multi-step reasoning, contextual understanding, and workflow execution across ML tasks. This agent can generate reports and insights as part of business processes.
• The Teradata monitoring agent provides automated monitoring and management of Teradata databases and infrastructure. By utilising real-time system telemetry, the agent can maintain system health, detect and respond to anomalies pre-emptively, and enhance performance across an enterprise's operations.

Availability

Teradata has stated that AgentBuilder will be released in private preview in the fourth quarter of 2025.