May Product Updates: Call Center AI

This month's updates for Call Center AI introduce smarter Speech Analytics, broader Voice Translation, and new ways to get your data where you need it. These enhancements help teams capture cleaner, more consistent call data with less manual effort, serve more customers in their own language, and bring call insights into the tools teams already use. Let's take a closer look!

Krisp Call Center AI updates

• Automatic call dispositions: Krisp now categorizes calls automatically in the After-call Summary, up to four levels deep. Agents close faster, and QA teams get consistent data without chasing it.

• Voice Translation selects the right language automatically: Krisp reads the call skill from your CCaaS and selects the Voice Translation language with no agent input and no mismatches.

• Voice Translation flags when someone new picks up: When a different person starts speaking on the customer side, agents get a quiet heads-up to re-verify or switch scripts, without the customer noticing.

• Broader language coverage: Voice Translation now supports Galician and Basque, with refreshed, more natural voices for Arabic, Catalan, German, Italian, Mandarin, Hindi, and Korean.

• Quick Phrases can now be personalized: An {{agent_name}} wildcard resolves to each agent's name during the call, so a single team-wide phrase greets every customer personally.

• Speech Analytics data, anywhere you need it: A new public API brings call scores, transcripts, and compliance signals into the BI tools and dashboards your team already uses.

• Smarter scoring with Salesforce context: Krisp pulls customer data from Salesforce before scoring, so scores reflect what actually happened on the call — steps the CRM already confirms are accounted for, and real gaps get flagged.

• Cleaner reporting by default: Krisp now ignores non-calling apps, so your dashboards reflect actual customer calls only.