Tell a regression from an infra failure
Preview text: Tell a regression from an infra failure 96 Scores, errors, and traces on one Harbor task…

Using this email in an article? Credit: Arize email via They Sent It, with a link to this page. Full-size image
- Subject
- Tell a regression from an infra failure
- Preview text
- Tell a regression from an infra failure 96 Scores, errors, and traces on one Harbor task…
- From
- mmeyer@
email.arize.com - Brand
- Arize
- ESP
- HubSpot
- Sent
- Tue Oct 6, 2026, 1:50 PM ET
- Sent to
- phoenix.arize.com
- Size
- 48 KB
- Images
- 5
- Links
- 5, 2 of them buttons
- Subject length
- 39 characters
Against Arize's usual
Sent in the 1 PM hour, which holds 18% of its emails; 18% arrive in the 9 AM hour. The subject is 39 characters; its average is 44.
Email headers
What the brand's servers put on this email. Our receiving servers, the address we subscribed with and unsubscribe links are left out.
- From
- Marine @ Arize <[email protected]>
- Reply-To
- [email protected]
- Subject
- Tell a regression from an infra failure
- Date
- Tue, 6 Oct 2026 10:50:03 -0700
- Message-ID domain
- 20083050m.email.arize.com
- Return-Path domain
- 20083050m.email.arize.com
- DKIM signature
- d=20083050m.email.arize.com s=hs1; d=email.arize.com s=hs2-20083050
- List-Unsubscribe
- Present, one-click
- Feedback-ID sender
- HubSpot
Other sending headers (names only): X-HS-Cid, X-Report-Abuse-To
Sent just before and after
| Nov 4, 2026 2:50 PM | An LLM judge on every span is the expensive path | |
| Nov 4, 2026 9:59 AM | Trace and evaluate your agent's decisions with Phoenix | |
| Oct 6, 2026 1:50 PM | Tell a regression from an infra failure (this email) | |
| Oct 6, 2026 11:54 AM | Debug agent failures from trace to code with Phoenix | |
| Sep 29, 2026 1:50 PM | A controlled agent benchmark, then the traces | |
| Sep 22, 2026 1:50 PM | Same tasks, different agents: what actually improved | |
| Sep 15, 2026 11:57 AM | Still time: durable agents you can actually evaluate |
More from Arize
All Arize emailsAn LLM judge on every span is the expensive path
Nov 4, 2026 2:50 PM
Trace and evaluate your agent's decisions with Phoenix
Nov 4, 2026 9:59 AM
Debug agent failures from trace to code with Phoenix
Oct 6, 2026 11:54 AM
A controlled agent benchmark, then the traces
Sep 29, 2026 1:50 PM



