ElevenLabs has added an internal AI assistant called Architect to ElevenAgents, its enterprise AI agent platform, according to BlockBeats citing Beating AI. The platform is designed for companies building AI systems that can answer phone calls, handle online chats, and take on customer service work.
Architect is built to help teams both create and maintain those agents. In one example, if refund-related calls are frequently handed off to human staff, a team can ask Architect why that is happening. The tool can review real call records, agent configuration, and test results, identify the issue, and then adjust prompts, workflows, tools, knowledge bases, or guardrails.
After making changes, Architect automatically writes simulated tests and compares performance before and after the update. Those edits do not go live immediately. They first remain in draft form, then move to human review and release after testing is complete. Architect can also set up new agents directly once users provide requirements and source materials. ElevenLabs said the tool is meant to support continuous improvement, though tasks still have to be initiated by humans and all changes still require human approval before deployment.
ElevenLabs has added an internal AI assistant called Architect to ElevenAgents, the company’s enterprise AI agent platform.
According to BlockBeats, citing Beating AI, companies use ElevenAgents to build AI systems that can answer phone calls, chat online, and handle customer service tasks. Architect is designed to help teams build and maintain those agents.
Architect reviews calls, configs, and test results
One example given was refund calls that are often escalated to human staff. Teams can ask Architect why that is happening. The tool can read real calls, agent configuration, and test results, identify the problem, and then modify prompts, workflows, tools, knowledge bases, or guardrails.
Once changes are made, Architect automatically writes simulated tests and compares performance before and after the update. All edits first go into draft mode rather than affecting live agents directly. After testing is completed, the changes are sent for human review and publication.
It can also build new agents
Architect can also create a new agent directly. After users explain their requirements and provide materials, it can configure prompts, knowledge bases, tools, and tests.
Teams can access Architect through Claude, Claude Code, ChatGPT, and Cursor. ElevenLabs said Architect can help agents keep improving over time. At this stage, though, tasks still need to be initiated by humans, and any changes must be approved by people before they can go live.
This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan. Disclaimer:
The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.
Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.
