Never let a signup go cold again.
Turn emails into revenue with Brew. No credit card, free credits to try.
Turn emails into revenue with Brew. No credit card, free credits to try.
inngest.com · newsletter
Explore this email design and adapt it to your own brand. Review the copy, links, and offer before sending.
If you're building agents in production, you've probably realized how painful (and expensive) it is to evaluate performance and outcomes. Especially if waiting for a key event to occur hours or days later.
Today, we're launching Agent Evals to help.
Score and improve agents just as easily as you make them durable—with APIs added to existing code. No separate instrumentation, no new pipelines. Inngest already has the data it needs—captured at execution.
What's Inside?
The following beta features comprise Agent Evals, and are available in v4.8.0 of the TS SDK.
Scoring — Attach a metric to any run or step, from inside or outside a function, including outcomes that don't arrive until long after the run finishes.
Sessions — Group like or connected runs and score together.
Experiments — Run variants of a prompt, model, or strategy against live production traffic and compare against execution data Inngest already records.
Defer — Launch durable follow-up work from inside a function, the primitive that lets a scorer run after a run completes and wait for what happens next.
Traces for AI — Capture model calls, token counts, and latency as span-level metadata on every agent run.
Note: All Inngest plans (including free!) will have unlimited access to these features until GA, when each plan will receive a baseline allocation that can be increased as needed.
Happy building,
Lauren
TwitterLinkedIn
About Terms Contact
©2026 Inngest
The email was sent to you@example.com. To no longer receive these emails, unsubscribe here.
Browse more email designs