IBM announced the new AI agent feature in its AI and data platform, watsonx.ai, last week. The feature offers AI builders a complete toolkit to streamline the AI development process at each step, as announced on the official blog.
The company’s CTO took to LinkedIn to announce the features for this latest development and suggested using watsonx.ai as an inference provider by using IBM and other popular open-source agentic frameworks.
“Easily build agentic services with watsonx.ai as an inference provider by using IBM and popular open-source agentic frameworks, such as LangChain, LangGraph and Crew AI through industry-standard API and SDK support,” Geeta Gurnan, CTO of Technical Community and Client Engineering at IBM Technology said.
This development follows IBM’s launch of the Bee Agent Framework on 26 October. The framework was designed to help developers build powerful agents with minimal changes to their existing setups.
IBM mentioned that this new feature helps consumers build, deploy and monitor AI agents in watsonx.ai.
AI agents help boost organisational productivity by managing complex business tasks autonomously. These AI agents need strategy and well-collected data to succeed. IBM claims to be focused on delivering secure, scalable, and reliable agent workflows.
Flexibility is an essential factor in AI agent development, as organisational requirements vary. The new feature will enable developers to build next-generation consumer and enterprise agent workflows, transforming how people interact with AI.
AI Services by Custom Agents
AI can streamline finance workflows and processes through automation and content generation. It typically handles complex, time-consuming tasks like data collection, document recap and summary, and pattern recognition.
IBM’s AI agents will help users easily deploy and monitor custom agent services and applications built in their own IDE.
Due to the framework-agnostic approach, developers can use watsonx.ai to either use IBM’s library of third-party and IBM Granite models or deploy and monitor their own agents or apps.
The IBM Low-code Agent Builder
According to the company, the AI agent builder speeds up development by easily managing agent services.
This IBM watsonx.ai agent builder is an lc-nc (low-code/no-code) tool, to be released soon, that will let developers create agents with pre-made workflows and an intuitive visual interface. It will be done by integrating with IBM watsonx Orchestrate to build custom agents, pre-built agents, and third-party frameworks together ‘powered with enterprise business logic and domain-driven design’.
Developers can now choose the agent’s architecture, select tools with a click, and design the agents with visual components that make setup simple.
Your reaction
Discussion
What's your take on this story?
No comments yet — be the first to share your view.