Observed arrival · 2026-10-19
Probative Labs checks what “resolved” really means
A browser-based audit compares AI support conversations marked resolved with later support tickets, using your own exports.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- IT and support teams evaluating AI agents
- Worth noticing
- It separates bot-opened incidents from failures and requires an Agent chat field to identify live-agent handoffs.
Field notes
The check accounts for platform-specific cases: a bot-opened incident is treated as a handoff, while catalog requests are not counted as returns. The page says it can show how results change under different baselines, including reopened tickets or every ticket counted. Founder Vivek Reddy Gaddam describes the project as a one-person effort and says the method is open source.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
$PaidCommerce or pricing visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue