Product updates ·
Gorgias reorganizes support analytics with AI and human comparisons
The October 13 update combines 14 reports into three and compares AI Agent with human support. The agent-type chart split excludes Auto-QA scores.
The industry, in context
AI agent & chatbot news · Page 5
What changed. What the sources say. Why it matters for your customer service team.
News feed (RSS, English)Latest news
Product updates ·
Cognigy adds timed Copilot sessions after conversation closure
The October 12 release keeps immediate closure as the default, but lets teams retain an Agent Copilot session for a configured period after a human resolves the conversation.
Product updates ·
Genesys announces prerequisite checks for virtual-agent tools
The October 12 release notes add knowledge-tool dependencies and input checks. Configuration documentation shows why collecting a variable and validating its value are different steps; availability follows a regional rollout.
Product updates ·
Parloa announces Context Intelligence for authenticated customer memory
The announcement connects retained customer facts with next-action recommendations. Support teams should evaluate who can retrieve that memory, how it changes and when selling is appropriate as separate decisions.
Product updates ·
Parloa announces Relationship Score with a ceiling for trust failures
The composite score puts conversation experience ahead of efficiency. A capped rating records a detected problem; it does not establish that the agent prevented it.
Product updates ·
ElevenAgents Architect enters alpha with distinct controls for drafts and live changes
The assistant can investigate support conversations and prepare fixes. Publishing a draft stays manual, but Auto-approve can allow other changes to reach live callers without another prompt.
Integrations ·
Ada adds role-based access for Amazon Connect chat handoffs
New setups can avoid sharing long-lived AWS keys. Recording which queue received a chat remains a separate, optional step.
Product updates ·
Kore.ai launches Autoloop with a choice of approval modes for agent changes
The Artemis optimization engine can apply fixes automatically or leave approval to people. Support teams still need to define the policies their tests must protect.
Industry ·
Automation Anywhere agrees to acquire Boost.ai; closing expected in Q4
The proposed combination links customer conversations with business-process automation. The deal announcement does not establish when customers could deploy a combined service.
Product updates ·
Sanas opens supervised AI calls by invitation; shared oversight needs testing
One supervisor can oversee up to four AI calls, Sanas says. That capacity claim leaves an operational question: what happens when several calls need a person at once?
Product updates ·
Parloa adds generated test drafts and version comparisons for support agents
Experts review the tests before use. The announced A/B comparison runs agent versions against matching scenarios; it is not evidence of better live customer outcomes.
Product updates ·
Sierra details fleming-1 voice detection; caller permission remains separate
Sierra says fleming-1 can flag likely AI callers for its voice agents. The company decides how to respond; the announcement provides no numerical error rates.
Showing 49–60 of 72 stories