Engineering time
Where does Ansible time actually go?
The repeat tickets, handoffs, and rework consuming the people you can least spare.
Your team knows how to automate it.
They don't have the time.
Your tickets sit in ServiceNow and Jira. Playbooks are still written by hand. Berch Emerson helps Ansible Automation Platform teams turn that backlog into reviewed, reusable job templates.
The problem
Licensing AAP doesn't write the playbooks.
Most infrastructure organizations already run the pieces: ServiceNow or Jira, Git, Red Hat Ansible Automation Platform, and engineers who know Ansible.
What sits between a ticket and a job template is still manual work. Someone has to read the request, write the playbook, test it, get it reviewed, commit it to Git, and wire it into Controller. That work runs on the same engineers handling today's requests. So the queue grows.
If this sounds familiar
- The same changes get made by hand every week.
- Tickets wait months for a playbook, then get done by hand anyway.
- Licensed AAP nodes run a fraction of the job templates they could.
- The engineers who write Ansible are the ones who can't be spared.
- Project work slips behind operational work.
If you have more tickets that should be playbooks than time to write them, you have an automation backlog.
The capacity gap
Where the work narrows.
Every request takes one of two paths. Done by hand, it costs engineering time every time it comes back. Built once into a playbook and AAP job template, it costs engineering time once. The constraint is writing the playbook — and it runs on the same people.
Every request takes one of two paths. Done by hand, it costs engineering time every time it comes back. Built once into a playbook and AAP job template, it costs engineering time once. The constraint is writing the playbook — and it runs on the same people. Today: Demand arrives faster than it can be built. Most work goes the manual route, the backlog grows, and engineering time is spent before project work starts. Constraint released: More work becomes repeatable. Each automation keeps running without new engineering time. The backlog shrinks, and time returns to the projects that were waiting.
Demand arrives faster than it can be built. Most work goes the manual route, the backlog grows, and engineering time is spent before project work starts.
Illustrative. No two organizations narrow in the same place. Finding where yours does is the work.
Where we look
Four questions that locate the constraint.
Backlog
What's waiting, and why?
Which ServiceNow and Jira requests repeat, how long they've aged, and which should become playbooks.
Ticket to job template
How long from request to a running job?
Where drafting, review, Git commit, and promotion to Controller slow the work down.
AAP return
Are your licensed nodes doing the work they should?
The true return on AAP is the job templates that exist and actually run.
How we start
Diagnose before prescribing.
The first engagement is a fixed-scope capacity assessment. It uses data you already have — ITSM ticket history, Git activity, AAP job records — and conversations with the people doing the work.
Observe
How work enters, where it accumulates, and how the team responds.
Measure
Where time, backlog, handoffs, and utilization limit throughput.
Prioritize
Which constraints carry the most operational and economic weight.
Recommend
What combination of process, standards, agentic AI, or tooling would turn more tickets into job templates.
What you receive
- A map of how demand enters and where it waits.
- A classified view of the ticket backlog: what should become a playbook, what shouldn't.
- The path from ticket to AAP job template, and where it stalls.
- A ranked list of constraints, with an estimate of the engineering time each consumes.
- A recommendation and a 90-day plan leadership can act on.
Afterward
Many teams carry the plan out themselves. Where it calls for tooling — for example, self-hosted agentic AI that drafts playbooks from tickets for your team to review — we recommend what fits, and disclose any commercial relationship before you act.
Technology doesn't create capacity.
Repeatable work does.
Not fewer engineers.
The point is to stop spending scarce engineering time on work that should already be repeatable, so it goes to work only your engineers can do.
Not AI for its own sake.
Agents can draft playbooks. People still review, commit, and approve what AAP runs — on infrastructure you control.
Not a recommendation before a diagnosis.
We recommend what removes the constraint. Sometimes that's a tool. Often it isn't.
About
Founder-led, by design.
Berch Emerson is a small firm focused on Ansible Automation Platform teams. We assess the automation backlog and recommend the process and tooling that clears it. The founder runs every engagement directly. There is no handoff.
Berch Emerson works commercially with a small number of technology companies. When a recommendation involves one of them, we disclose it before you act on it.
Do you have more work to automate than time to automate it?
Start with a conversation about where your work is waiting.
Start a conversation