Skip to content

Customer support & contact centers

Amazon Connect reporting on contacts, queues and agents in your own words

Amazon Connect Customer, previously called Amazon Connect, is the AWS cloud contact centre service. Operations teams configure instances with queues, routing profiles, flows and hours of operation, and agents handle voice, chat and task contacts in the Contact Control Panel. Every contact produces a contact record, and Connect keeps real-time and historical metrics for queues, agents and routing profiles.

Amazon Connect

Available now: Nexatron answers Amazon Connect questions with user-level access that inherits your existing Amazon Connect permissions.

Updated

Can Claude, ChatGPT or Copilot access Amazon Connect data?

Yes. To connect Claude, ChatGPT or Copilot to Amazon Connect, add Nexatron from Claude’s connector directory, or by its MCP server URL in ChatGPT and Copilot. Then connect it in Nexatron. Users sign in with their own Amazon Connect account (or your SSO), and Nexatron inherits their existing permissions.

What Amazon Connect data a question reads

Contact records (CTRs)
One record per voice, chat or task contact, with channel, initiation method, queue, agent, initiation and disconnect timestamps, hold and after-contact work durations, attributes set in the contact flow and disconnect reason.
Queues
Standard and agent queues with their hours of operation, outbound caller ID and maximum contacts, which is where service level and abandonment are measured.
Routing profiles
The queues and channels each agent can take, with priority and delay settings and the concurrency allowed per channel.
Users and agent status
Agent accounts with security profile, routing profile and hierarchy group, plus agent status history such as Available, Offline and custom break states.
Historical metrics (GetMetricDataV2)
Aggregated metrics by queue, routing profile, agent or hierarchy over an interval, such as contacts handled, average handle time, service level, abandonment rate and agent occupancy.
Contact Lens analysis
Where Contact Lens is enabled: transcripts, sentiment scores, detected issues, categories matched by rules and talk and non-talk time per contact.

Amazon Connect questions people ask

  • What was service level at 20 seconds for each queue last week, by day and hour?
  • Which routing profiles had the highest average after-contact work time this month?
  • How many contacts were abandoned in queue before reaching an agent yesterday, and how long had they waited?
  • Which agents in the Tier 2 hierarchy group had occupancy above 90 percent during peak hours?
  • How did average handle time for chat compare with voice in the billing queue over the last quarter?
  • Which Contact Lens categories appeared most often on contacts that were transferred more than once?
  • What share of inbound calls on the sales line were handled after hours of operation ended?

Who asks Amazon Connect questions

Contact centre operations
Service level, abandonment and queue backlog by interval, to see where staffing fell short of demand.
Team leaders and supervisors
Handle time, after-contact work and status time for the agents in their hierarchy group.
Customer experience analysts
Contact Lens categories, sentiment and repeat contacts, linked to the queues and flows that produced them.

How Nexatron connects to Amazon Connect

API and sign-in
The Amazon Connect API (GetMetricDataV2, GetCurrentMetricData, SearchContacts, DescribeContact) and contact records delivered through Kinesis to S3. Calls are signed with AWS Signature Version 4 using IAM credentials, which IAM Identity Center can issue to each user for their permission set.
User-level access
Users sign in with their own Amazon Connect account (or your SSO), and Nexatron inherits their existing permissions. Users only see data they are already authorized to see.
Read-only
Questions only read Amazon Connect; asking a question never changes Amazon Connect data.
Setup
An administrator enables Amazon Connect in Nexatron. Users link their own Amazon Connect account, or sign in through your SSO, the first time they ask. Queries run in Amazon Connect, with no warehouse to load.

Amazon Connect in Claude’s connector directory

Nexatron is listed in Claude’s MCP connector directory, so a Claude user can add it without a custom server URL. Its hosted MCP server answers Amazon Connect questions for Claude, ChatGPT and Microsoft Copilot with user-level access, and the same governed answer as the Nexatron app. It is Nexatron’s own server, not one published by the vendor.

  1. Connect Amazon Connect to your Nexatron workspace.
  2. In Claude, add Nexatron from Claude’s connector directory. In ChatGPT or Copilot, add a custom MCP connector with Nexatron’s server URL.
  3. Sign in with your own Nexatron identity (OAuth 2.1), then ask your Amazon Connect questions.

Server URL for a custom connector: https://api.nexatron.io/api/v1/mcp

Amazon Connect questions and answers

Where does Amazon Connect store historical reporting data?
The GetMetricDataV2 API can retrieve historical data for the last 3 months, and contact records are retained for 24 months. Teams that need longer history or custom joins stream contact records and agent events through Kinesis into S3 and query them with Athena or a warehouse.
What is the difference between real-time and historical metrics in Amazon Connect?
Real-time metrics show what is happening now: contacts in queue, oldest contact age and agents by status. Historical metrics are aggregated over a closed interval, such as service level, contacts handled and average handle time per queue or agent. Questions about yesterday or last month belong to the historical set.
Can Amazon Connect reporting include Contact Lens data?
Yes, when Contact Lens conversational analytics is enabled on the contact flow. It adds transcripts, sentiment, categories and talk-time measures per contact, which can be filtered next to the usual queue metrics. Nexatron treats these as contact attributes and only shows them to people whose IAM access covers that analysis output.
How is service level calculated in Amazon Connect?
Service level is the percentage of contacts answered within a threshold you choose, such as 20 or 60 seconds, out of contacts that entered the queue. GetMetricDataV2 lets you set the threshold per request, so a question should name the threshold and whether abandoned contacts count in the denominator.