AI-Native ClubBlog

London / Global

A practical guide / Facilities management

Can AI work with your CAFM system?

Facilities management teams can use AI to turn reactive job notes into a clear log. It can also sort planned visits and draft contract reports from job data. Start with the work order and asset register, then shape the output for the site team or client.

A good first project is a monthly contract report. The structure below connects finished work, repeat faults and open actions. The contract manager can use it to explain the position and agree priorities.

Prepare a contract report
Connect site jobs, asset history and client decisions.
Illustration: a building plant room of paper-craft pipes and panels with one orange maintenance tag on a valve.

Connect site jobs, asset history and client decisions.

  1. Make reactive notes useful beyond the helpdesk

    Use one record for each reported issue. Capture the job ID, site, asset tag and reported symptom. Add the visit, work finished and next action. Ask AI to turn free-text notes into these fields while keeping the original note ID. Keep reported symptoms separate from the engineer’s findings.

    More on this step

    Group repeat calls by checked asset tag and issue type. This gives the manager a way to review a repeat fault across several visits. Show the individual job IDs under each group so the engineer can follow the history.

    Use the same log to prepare an end-of-day handover. List jobs awaiting parts, access, specialist visit or client agreement. Include the owner and next update date. A colleague covering the desk can then explain what each job needs without reading the full message trail.

  2. Build the maintenance schedule from agreed tasks

    Gather the asset register, approved maintenance tasks, contract intervals and planned access windows. Use AI to sort these into a calendar. Give each entry an asset and task ID. Add the interval, due date, provider and access needs. Keep the source of the interval visible.

    More on this step

    Use calendar rules to set dates from the approved schedule. Ask AI to summarise clashes, group visits by site and draft visit-planning notes. Give contractors the asset list and task IDs for each visit. Ask them to check resources and timing.

    Track planned, booked, visited and finished as distinct stages. Keep any remaining defect linked to the original visit. Use these job records in the contract report. Show the work delivered and follow-up still due.

  3. A monthly report the client can act on

    Reuse these headings: service overview, reactive work, planned work, repeat issues, costs against approved figures and decisions required. Ask AI to draft a short explanation for each, using the period’s job export. Put the report dates and source export version at the top.

    More on this step

    Calculate totals in the job system or spreadsheet and supply those figures with the brief. Use the terms in the contract when drafting the text. Define “finished” in your reporting process. Does it mean an engineer visit or accepted close-out?

    Finish with a decision list. Each item should name the asset, current issue, planned action from the manager, cost ID and date needed. The client ca reply against each item. Carry agreed decisions into the next report.

Worked example

From three call-outs to one clear issue summary

Example with made-up details: the log for fan F12 contains three call-outs in a month. The first two record a reset. The third records a replacement part ordered, with access for a return visit still to be confirmed.

The report entry reads: “F12 generated three reactive jobs this month. The latest engineer note records a replacement part order. The helpdesk is sorting access for the return visit.” The entry links all three work orders and names the helpdesk coordinator as the next-action owner.

The manager can now discuss the repeat fault with the client and sort the follow-up. The report shows visits already made. It also shows the work still due.

Illustrative example.

How this example works
  1. Job records

    Organise work by asset and status.

  2. Contract report

    Explain service delivery and recurring issues.

  3. Next decisions

    Assign follow-up and client actions.

Use competent approval and secure site records

HSE says work equipment maintenance should be carried out by competent people. Its inspection guidance explains that inspection needs depend on equipment and its risks. Use competent review to set tasks, inspection scope and service intervals. Apply the requirements for each asset.

The ICO’s AI guidance addresses personal data security. Restrict access to staff and contractor details in site records. Keep access codes and sensitive security arrangements in the approved site system. The contract manager approves the report; the right technical person confirms the work and asset status.

Read the source: HSE: maintenance of work equipment; HSE: inspection of work equipment; ICO: guidance on AI and data protection

Try a practical task

Draft our monthly contract report

Prepare a report for [site] covering [period] from this job export, approved maintenance schedule and verified totals. Use our contract terms for close-out and response categories.

Try this in your AI app

Using Gemini? The button copies the prompt. Paste it into the new chat.

Write a short overview, then cover reactive work, planned visits, repeat issues and outstanding actions. Link each issue to the relevant asset and job IDs. Keep logged statuses. Distinguish a booked visit from finished work.

Use the supplied cost and service figures. For each client decision, give the source and manager’s recommendation. Add the date a response is needed. End with a list of missing records to gather before issue. Keep the full job history in an appendix and the main report concise.

A relevant public story

Iain connected audit evidence to reporting

Iain Davidson audits health, safety and security at luxury hotels. With coaching, he built a Python app for evidence, issue scores and reports. He produced a 70-page first client report and used the tool at further European hotels.

His work shows how a facilities team can connect site evidence to the information a client needs to act.

Read the source: Remote Humans: Iain’s hotel audit system

Testimonial · 4:16

Speed

"It's going to save a considerable amount of time. More importantly, it's going to save a considerable amount of money."

Iain Davidson · Health, Safety & Security Leader, Five-Star Luxury Hotels

Leader Lab

Questions about this task

Can AI set our maintenance intervals?

Use the intervals agreed for each asset through the relevant technical and contract process. AI can sort those tasks into a schedule, group visits and draft the access requests.

What data do we need for a useful contract report?

Begin with job IDs, asset tags, dates, logged work, statuses and next actions. Add approved service terms and verified totals. These give the narrative a clear link to the work delivered.