Most QA questions are one-off: which agents dropped below 70 last week, how many fatal incidents in collections this month, which work code is growing. Building a dashboard for each is not worth it. The AI Assistant answers them directly.
Agent rankings and per-agent score detail. Category and parameter-level breakdowns. Fatal incident reports and counts by agent. Compliance rates for any individual quality parameter. Call volume and duration trends by date. Keyword frequency, angry-keyword analysis and free-text search across call content. Silence and dead-conversation analysis. Work code breakdowns. Daily and monthly usage, and processing cost.
The assistant enforces the same role and data scope as the rest of the product. A team lead scoped to one department sees only that department; an agent-scoped user sees only their own calls. The restriction is applied inside the query itself, not filtered afterwards.
Upload your QA policy, scripts or product documentation and the assistant searches them alongside call data. Tables and informative images inside those documents are summarised so they can be retrieved as text, not skipped.
The assistant reaches your data through a fixed set of read-only operations against approved fields — it cannot write, delete, or reach another tenant's data, regardless of what it is asked.
See it working
Questions
It answers natural-language questions about your call data — agent rankings, category and parameter scores, fatal incident reports, compliance rates, keyword and sentiment analysis, silence detection, call volumes, work code breakdowns, and usage and cost — returning results as tables and summaries.
No. It applies the same role-based scope as the rest of Xperia, enforced inside the query rather than filtered afterwards, so a department-scoped or agent-scoped user only ever receives their own data.
Yes. Uploaded documents in PDF, Word, Excel, CSV, text and Markdown are indexed alongside call data, including the contents of tables and informative images, so the assistant can answer from your QA policy or scripts as well as your calls.
No, it is a separate module at $100 per month or $1,000 per year, on top of per-minute call analysis.
Related
Xperia scores every call against 48 quality dimensions across 10 categories. Weightings, pass criteria and the parameters themselves are configurable per department.
Xperia analyses call recordings in over 50 languages, including English, Urdu, Arabic, Hindi, French, Spanish, Bengali and Punjabi, and handles mid-conversation language switching.
Every account starts with $3 of free credit — around 75 minutes of call analysis. No card required.
Start Free