Agent assist gives agents real-time recommendations and relevant content while they're actively on a contact. It automatically detects customer intent during calls, chats, and emails using conversational analytics and natural language understanding (NLU), then provides immediate, generative responses and suggested actions. [Source: AWS, Use Amazon Q in Connect for generative AI–powered agent assistance in real-time]
What It Actually Does, Moment to Moment
As a conversation unfolds, agent assist listens or reads along, works out what the customer is actually asking for, and surfaces the information an agent would otherwise have had to search for manually — knowledge base answers, next-best-action suggestions, and (depending on configuration) drafted responses or post-contact summaries. [Source: GDC IT Solutions, Amazon Connect Agent Assist for Contact Centers]
Channel Coverage
| Channel | How recommendations appear |
|---|---|
| Chat | Recommendations generated and shown by default |
| Recommendations generated and shown by default | |
| Voice | Recommendations appear if conversational analytics is enabled, since that's what provides the live call transcription agent assist works from |
[Source: AWS, Use real-time recommendations from Connect Customer agent assist]
Personalisation From Customer Data
Agent assist doesn't operate in isolation from the rest of Amazon Connect — one of its stronger capabilities is drawing on customer data to make recommendations highly personalised and context-aware, rather than generic knowledge-base answers that ignore who the specific customer is and what they've already told the agent. [Source: AWS, Personalize agent assistance with Amazon Q in Connect] This is where its integration with Customer Profiles matters in practice.
Built on AI Agents, Not a Separate Bolt-On
Agent assist is powered by configurable AI agents within Amazon Connect — specifically an Answer recommendation AI agent that drives the automatic, intent-based recommendations pushed to the agent during a contact. [Source: AWS, Create AI agents in Connect Customer] These AI agents make proactive recommendations based on the real-time interaction, and help guide the human agent toward the right resolution path. [Source: AWS, Use Connect Customer agentic assistance]
Where this came from: this capability traces back to Amazon Connect Wisdom, which let agents search a connected knowledge repository and get real-time recommendations from the agent desktop, with built-in connectors to systems like Salesforce and ServiceNow. [Source: AWS, Ingesting content to power real-time recommendations and search with Amazon Connect Wisdom] The current generative, AI-agent-powered version is the evolution of that same idea.
Agent Assist vs. a Self-Service AI Agent
It's worth being precise about the distinction: agent assist supports a human agent during a live interaction. A self-service AI agent, by contrast, engages the end customer directly over voice or chat and can resolve issues autonomously, with no human agent involved at all unless it escalates. [Source: AWS, Use AI agents for real-time assistance] They're complementary capabilities within the same platform, not the same feature under two names.
Frequently Asked Questions
What is Amazon Connect agent assist?
Agent assist is the Amazon Connect capability that gives agents real-time recommendations and relevant content while they're on a live call, chat, or email. It automatically detects customer intent using conversational analytics and natural language understanding, then surfaces generative, context-aware responses and suggested next actions.
What powers Amazon Connect agent assist?
Agent assist is powered by AI agents within Amazon Connect, including an Answer recommendation AI agent that drives automatic, intent-based recommendations pushed to the agent during a contact, drawing on connected knowledge sources and customer data.
Does agent assist work on calls as well as chat and email?
Recommendations appear by default for chats and emails. For voice calls, real-time recommendations require conversational analytics to be enabled, since that is what provides the live transcription agent assist analyses.
Is agent assist the same as a self-service AI agent?
No. Agent assist supports a human agent during a live interaction with recommendations and information. A self-service AI agent, by contrast, engages directly with the end customer over voice or chat and can resolve issues autonomously without a human agent in the loop.
Conclusion
Agent assist is the AI layer standing just behind the human agent — reading the conversation, understanding intent, and surfacing the right information before the agent has to ask for it, personalised using the same Customer Profiles data available elsewhere in the agent workspace. It's one of the most direct ways AI shows up in day-to-day agent work without replacing the agent at all.