Singlife’s Salesforce Agents Cut Handling Time 30% — And Halved New-Hire Training
Buddy on the desk, email agent on the queue: AHT −30%, training 4 weeks → 2, GI Email Agent >20% of enquiries. ASEAN insurer from pilot to operating KPIs.
Singlife (Singapore insurer) reports production metrics for Salesforce AI agents: Buddy for employee assist, plus a General Insurance Email Agent on the queue. This is pilot-exit territory — operating KPIs, not a demo reel.
What the numbers say
- Buddy: average handling time −30%
- New CS hire training: 4 weeks → 2 weeks (−50%)
- GI Email Agent handles >20% of email enquiries
- Most emails resolved within 2 days during a volume spike
Operator takeaway
Separate the task map: routine email triage → augment/automate; complex policy judgment → leave with humans who now have more time. Training time dropping in half is the quiet win — onboarding is a process, not a headcount line. Same BuildBrain leave/augment/automate lens: put agents on the repeatable slice, keep people on exceptions.
Curious where your CS email and onboarding tasks sit on that map? A free Ada report can inventory them without a paid audit first.
Caveat: metrics are from a TechNode.Global report on Singlife’s Salesforce deployment. Volume, seasonality, and definition of “handled” matter — verify before you forecast headcount.
Sources: TechNode Global — Singlife Salesforce AI agents handling time and training results.
