Monitoring agent · Design-partner pilot

Point it at one queue manager and watch.

A 30-day pilot of the Abendly monitoring agent: read-only, on one queue manager you choose, in your environment, with your own AI key. Founder-led, so you work with the person who built it. You see what it catches on your own estate, and the pilot report is yours to keep whether or not you continue.

Apply for the pilot → See the console first
FeeThe pilot has a fixed fee, credited in full toward your first year if you continue. It is quoted with the scope before you commit. How pricing works
Scope

Small on purpose.

One queue manager is enough to see the agent learn its baselines and catch real problems while they build, without a big security review up front.

Scope
1 queue managerYou choose it: distributed or z/OS, production or pre-production.
Access
Read-onlyInquire and display authority only, enforced by OAM or RACF. You provision it; you can revoke it.
Where it runs
Your environmentYour cloud account or VPC, or your data centre. Details agreed at kickoff.
AI model
Your keyAnthropic, Azure OpenAI, Amazon Bedrock, IBM watsonx, or a self-hosted model. Identifiers are replaced with labels before anything is reasoned over.
Delivery
Console and SlackFindings arrive in the console and, through a Slack webhook you create, in your own Slack workspace. PagerDuty delivery is on the roadmap, not in the pilot.
Length
30 daysBaselines first, then live warnings, then a written report.
Who
Founder-ledA design-partner cohort, not a self-serve trial.
service account · scope
  • DISPLAY CHSTATUS
  • DISPLAY QLOCAL / QDEPTH
  • read SYSTEM.ADMIN.*.EVENT
  • STOP / START CHANNEL
  • ALTER / DEFINE / DELETE
  • PUT / GET / BROWSE application messages
Enforced by OAM on distributed and RACF on z/OS. You provision it, and you can revoke it. We send the exact definitions before Day 0.
30 days

Connect, learn, warn, report.

The bar is to scale: a week of learning what normal looks like, then three weeks of live warnings.

  1. Day 0

    Connect, read-only

    Your team applies the inquire and display definitions we send, and the agent connects to the queue manager you chose. Nothing else changes.

  2. Days 1–7

    Learn the baselines

    The agent learns what normal looks like for each queue and channel: depth ranges, rise rates, quiet hours.

  3. Days 7–30

    Live warnings

    Findings arrive in the console and in your Slack workspace, with the cause, a confidence level, the runway to impact and the MQSC fix in execution order. Your engineers decide what to run.

  4. Day 30

    The pilot report

    What it measured on your queue manager:

    • Warnings issued before anything failed
    • Median runway: how early the warnings came
    • Engineer-verified accuracy: findings your team confirmed
    • Pages avoided: warnings resolved before an existing alert fired
    • Resource use, measured: CPU on the collector host and on your queue manager (on z/OS, CPU seconds and MSU), against the published design targets
Both sides

What we need, and what you get.

What we need from you

  • One queue manager to point it at, and someone who can approve that.
  • A read-only service account, set up from the definitions we send.
  • Somewhere to run the agent in your environment, agreed at kickoff.
  • Your AI provider key, or a self-hosted model endpoint.
  • A Slack incoming webhook into your workspace, for the channel findings should go to.
  • An engineer's time to confirm or reject findings during the pilot, so the accuracy figure is yours, not ours.

What you get

  • Live findings on your own queue manager, in the console you can see in the live demo and in your Slack workspace. PagerDuty delivery is on the roadmap.
  • Direct access to the founder for the whole pilot.
  • The Day-30 report, yours to keep even if you don't continue.
  • The pilot fee back as credit toward your first year, if you continue.
Apply

Join the design-partner cohort.

Tell us a little about your estate. We'll reply within one business day to set up a short call about scope and fit.

  • A fixed scope and a fixed fee, agreed before you commit
  • The pilot has a fixed fee, credited in full toward your first year if you continue
  • Not ready yet? Start with a health check

We use these details only to reply about the pilot. No marketing list. Please don't include queue manager names or hostnames here. Privacy policy

Questions

Asked plainly.

Is the monitoring agent generally available?
Not yet. Today the monitoring agent is available as a design-partner pilot on one queue manager. General availability at full estate scale is on the roadmap. The log analyser and the one-time MQ health check are available now.
Is the pilot free?
No. The pilot has a fixed fee, credited in full toward your first year if you continue. The pilot report is yours to keep either way.
Can the agent change anything on our queue manager?
No. It runs under a service account with inquire and display authority only, enforced by OAM on distributed and RACF on z/OS. You provision that account and you can revoke it at any time. Every fix it recommends is for your engineers to review and run.
Can we start on a non-production queue manager?
Yes. You choose the queue manager. A busy production queue manager shows more in 30 days, but a pre-production one is a reasonable place to start while your security review completes.
What happens at the end of the 30 days?
We walk through the pilot report with your team. If you continue, the subscription is quoted for the estate you want watched and the pilot fee comes off the first year. If you don't, you revoke the service account and keep the report.