Apply · The Governed Workflow · Finance and operations teams
Learn how to turn a recurring Finance task into a governed AI workflow.
In this live 2.5-hour workshop, your team learns one method for using AI in Finance work, worked through together on a standard example: where AI belongs, what must be computed in code so the answer never changes, and what a person reviews before the result is relied on. Participants then take a second example through the same steps in pairs.
Learn one method. Apply it to many Finance tasks.
AI handles defined preparation. A person reviews. Your controls remain in place.
2.5-hour workshop · up to 8 · your office or online · $4,500
The course
About this course
Most Finance teams already have people using AI on their own. What they lack is a shared way to decide where it belongs, what it may touch, and who checks the result. This workshop gives them one.
The whole room works through one standard Finance example together, an AP sub-ledger to GL reconciliation on practice data, so everyone learns the same steps in the same order: what goes in, which steps stay deterministic (computed in code, the same answer every time), what AI prepares, what a person reviews, and how to tell when something has gone wrong.
In the second half, participants work in pairs on a second practice example, a different kind of recurring task, and take it through the same one-page 1113 Workflow Card—a structured way to examine one recurring task, while the instructor reviews. Practising that transfer on safe data is what lets the method carry over to the team’s real recurring work after the session.
The workshop runs on whichever AI tool your organization allows: Claude, Copilot, ChatGPT, or another. Tools change; the method carries over.
- Who it is for
- Finance and operations staff: accounting, FP&A, AP and AR, finance systems, and the ops, HR, and procurement teams around them
- Format
- Live 2.5-hour workshop. The team works through one standard Finance example together, then takes a second practice example through the same method in pairs. Practice data is provided.
The 1113 Workflow Card
Seven questions, asked of every task.
Every task goes through the same seven questions, so a team can compare one task with another and a reviewer knows exactly what to look for. A card-sized task has one recognizable trigger, one main input set, one main output, and one person or role responsible for reviewing the result.
- 01Purpose and trigger — Why does the task exist, and what starts it?
- 02Inputs and permissions — What does it use, and may that information enter the approved AI environment?
- 03Current steps — What happens today, in order?
- 04Rules, judgment, and AI’s role — What must be exact, what needs judgment, and what does AI prepare?
- 05Output and downstream use — What is produced, who uses it, and what happens next?
- 06Checks, exceptions, and responsibility — How will errors be found, and who remains accountable?
- 07Evidence and next decision — What did the test show: continue, revise, decompose, or stop?
What participants learn
Outcomes
- Identify recurring Finance work that may be suitable for AI assistance
- Define a task at a workable size
- Separate deterministic rules, human judgment, and an appropriate role for AI
- Document inputs, permissions, outputs, checks, exceptions, and responsibility
- Conduct a bounded test using approved or practice data
- Evaluate the result and decide to continue, revise, decompose, or stop
- Reuse the same method when considering additional tasks
Outline · 2.5 hours
How the session runs.
Usually 9:30 to 12:00 or 2:00 to 4:30. Other times are fine.
- 0120 min
Where AI helps in Finance work, and where it does not
- What these tools do reliably, and where they guess.
- What deterministic means (same input, same answer, every time, because the math is computed in code), and why the tie-out never goes to AI.
- The pattern behind every governed workflow: a defined input, a bounded transformation, a defined output, and a person who reviews it.
- Which data may be used, answered from the guide sent ahead.
- 0240 min
The method, worked through together
- One standard Finance example, an AP sub-ledger to GL reconciliation on practice data, taken through all seven fields of the Workflow Card with every screen in the same state.
- AI prepares only what the card allows: the explanations and the approval email. The tie-out stays a formula.
- 0330 min
Testing it, and catching the error
- Run it in the AI tool your organization allows.
- A planted error, and the ten-second check that catches it. Reviewing AI output with professional skepticism: against the source, never against how confident it sounds.
- The fix goes on the Workflow Card, where the next person will see it.
- Pairs review each other’s card for missing steps and unclear checks.
- 0420 min
The control layer
- Which data stays out, and which organizational policies apply.
- Who reviews the result, what evidence they need, and what happens when something is wrong.
- Three examples side by side: one ready to test, one that needs work, and one that should not use AI at all.
- 0540 min
A second example, in pairs
- In pairs, participants take a second practice task, a different kind of recurring Finance work, through all seven fields of the Workflow Card.
- The instructor reviews each pair’s card: whether the task needs to be decomposed, which steps stay deterministic, and what a person must review.
- What sits around a governed AI session: the card, the saved instructions the tool reuses (Claude calls them skills), the reference files, and the log.
- How the same method applies to the team’s own recurring work after the session.
Bring
What to bring
- A laptop with access to whichever AI tool your organization allows. Setup instructions arrive ahead of the course.
- Nothing needs to be prepared or sent in advance. Practice data is provided.
If IT blocks the tool, tell us on the call and we plan around it.
About your data
About your data
Courses use practice files that resemble real Finance work, so confidential data is not required to learn the method. A participant may use a sanitized file—or a company file only when the organization’s policy and approved AI environment permit it.
The method makes the input, transformation, expected output, and human review explicit. It does not require an unrestricted agent inside company systems, and it does not replace the organization’s technology, privacy, security, records-management, data-residency, or approval requirements.
Where a workflow could change a system of record, nothing posts without authorized human approval.
Instructor
Neil Perejma
Founder, 1113 Technologies and EPM ONE Consulting
Neil Perejma has advised Finance and operations leaders for more than 20 years on the processes, reporting, and controls their organizations depend on, and every engagement has left the team able to carry the work forward.
Today his clients are asking how to bring AI into Finance work without weakening governance. His courses give Finance teams a practical way to do it, keeping AI inside the controls they already have.
Workflow examples
Recurring Finance work this course is built around.
The others
Other courses
Build · Process owners and finance systems
Build a Governed Agent
Learn how to take a suitable process from a bounded workflow design to a scheduled, logged agent with deterministic rules and a human approval gate. The group builds and tests a working workshop prototype, then documents what would be required before organizational use.
Lead · CFO, controller, finance leadership
AI for the Finance Leadership Team
Learn how to evaluate bolt-on versus structural change, compare buy/build/both options, define a control model, and draft a 90-day action plan. The leadership team works through the decisions; the course does not make or implement those decisions for the organization.
Advance · Advanced · finance systems teams
Claude Code for Finance Systems
Advanced training for Finance-systems teams that already script. Participants learn and practise the conventions for building, testing, reviewing, documenting, and governing Claude Code agents, using guided and independent workshop exercises.
Questions about training your team?
Book a 20-minute call with Neil Perejma, on Zoom or by phone.