Use caseQuality assurance

Every case reviewed against your standards.

Agents review every case, human or agent, against your policies and procedures, flag what broke, and route the fix to the right owner so the same issue does not repeat.

How it works

From a closed case to a fix that sticks.

  1. 01

    Comes in

    Every closed case is picked up for review, whether a person or an AI agent handled it, across chat, email, calls and back-office work.

    /review
    └ 412 cases closed this week
    › 268 by people · 144 by agents
    › scorecard: complaints v4
    ✦ 412 of 412 queued
  2. 02

    Worked across teams

    Each case is scored against your scorecard. Findings go to the team lead, the policy owner or the agent builder, with the evidence.

    /findings
    › complaint not logged · 9 cases
    › refund reference missing · 37
    › routed to complaints team lead
    ‖ procedure change · awaiting owner
  3. 03

    Closed with evidence

    Fixes are tracked until they land, and the next week’s reviews show whether the issue stopped.

    /verify
    ✓ procedure updated
    ✓ 37 cases corrected
    ✓ issue rate back under target
    ✓ audit pack exported

Systems

Reads every channel your cases come through.

Agents read conversations and case records where they already live, and send findings to the people who own the fix.

Customer channels
  • Zendesk
  • Intercom
  • Gmail
  • Twilio
  • Call recordings
CRM & data
  • Salesforce
  • HubSpot
  • Snowflake
Policies
  • Google Drive
  • Notion
  • Procedure documents
Teams
  • Slack
  • Jira

Security and governance
for every agent action.

Agents act only within the permissions you set, pause for approval on high-risk steps and log every action with its evidence. No AI training on your data.

Read more
/scope
✓ read: ledger, CRM, inbox
✓ act: email, refund ≤ £500
✕ cannot: close accounts

Scoped permissions

Decide which systems, actions and limits each agent can use. Nothing happens outside that scope.

/approval
› step needs sign-off
‖ paused · routed to ops lead
✓ approved 10:42

Human approval

Set approval points for high-risk steps. Agents pause and route the decision to the right person.

/audit
09:14 message sent
10:42 step approved
11:05 evidence attached ✓

Full audit trail

Every action is logged with what was done, why, and the evidence behind it, ready for audit.

Certified security

SOC 2 Type II, GDPR and PCI DSS. AES-256 encryption at rest and TLS in transit.

More use cases

See every case reviewed against your standard.

Necessary

Enables core site behavior and stores your privacy choice in this browser for 180 days.

Always on

PostHog measures page views and interactions. It uses in-memory identifiers, and session recording is disabled.

Mesh by Avina records session and form interactions and can use device or IP data to resolve business visitors through providers such as Vector or RB2B.