Introducing Support QA: Enterpret now grades every support conversation, human and AI

Today we're launching Support QA. Enterpret now grades every support conversation on whether the customer's problem was actually resolved, judged from the transcript and the customer's own words instead of the ticket status. It scores your agents, human and AI, against your own rubric, sends each manager a coaching brief with the conversations cited, and routes failures caused by the product to engineering.

Enterpret already reads your support conversations to tell you what customers are asking for. What it didn't tell you was how well each conversation went. A helpdesk marks a ticket closed when it stops moving, and an AI agent vendor reports its own resolution rate. Neither tells you whether the customer walked away helped. Support QA adds that grade and keeps the reason next to it, so you know whether a failure calls for coaching, a process change, or a product fix.

What Support QA does

Grades whether the customer was actually helped. True Resolution reads each conversation and places it in one of six states: resolved, unresolved, escalated, agent stalled, customer dropped, or unknown. The grade comes from the transcript and what the customer said, independent of the status your helpdesk or AI agent vendor reports. Put Enterpret's resolution rate next to the one in your support platform for the same conversations and you can see exactly where the two disagree.

Shows how the customer felt by the end. Sentiment Shift compares the customer's tone at the start of a conversation with their tone at the end. You can see where the experience improved, where frustration held, and where it escalated, then click through to the conversations behind every number. Both signals live in Quantify and dashboards, and Enterpret Agent can build a dashboard on them for you.

Scores conversations against your rubric. Upload your existing QA rubric to Enterpret Agent as a skill, or describe what you care about and work with Enterpret Agent to write one. Run it as an automation on a schedule, scoped to a team or an individual agent, and it returns a scorecard with the conversation evidence behind each score.

Turns findings into coaching. On the cadence you set, each support manager gets a coaching brief in Slack or email. It lists the recurring behaviors, why each one matters, what to coach, a reply template, and the conversations that show it. A brief can cover a whole team or one agent, ready to share with the team or bring to a one-on-one.

Grades AI agents on the same scale as people. Every agent that handles a conversation, human or AI, becomes a support agent in your customer context graph with its own dashboard. Resolution and sentiment shift roll up by agent, team, and BPO, week over week. Your AI agent gets an independent audit from a party with no stake in the answer, measured the same way as your team.

Sends product failures to the team that fixes them. Every graded conversation carries the same taxonomy as the rest of your feedback, so a failed conversation arrives with its contact reason attached. When the cause is the product rather than the agent, an automation files it in Jira or Linear with the conversations cited.

Why we built it

Most QA teams can read only a small share of their conversations, and the resolution numbers that get reported upward come from systems with every reason to look good: a ticket marked closed, a bot vendor's own report. As more support runs through AI agents and outsourced teams, that gap gets harder to see and harder to argue about.

The part we care about most is the split between agent and product. When a conversation fails, a support leader needs to know whether to coach someone or escalate a broken flow. Support QA builds on the support conversations Enterpret already understands, so the grade, the reason, and the evidence sit in one place and the next step is clear.

Get started

To learn more about Support QA, reach out to your Enterpret CSM. Support QA currently supports Intercom, Zendesk Support, and Salesforce Service Cloud.

Your CSM will help you bring in your existing rubric, set the coaching cadence, and choose where coaching briefs and product tickets get delivered. The help center has the details.

If you're not using Enterpret yet, book a demo to see how Support QA works on your own support conversations.