AI Job Report

The AI Job Report is generated automatically at the end of every closed job. It captures everything that happened in a single, readable document — and it’s also exposed to the customer in a friendlier version through their portal.

​

When the report generates

The report is generated:

• At inspection complete — the moment you click Complete Inspection at the end of Return & Inspect
• From the closed job’s Overview tab — click Regenerate Report any time after, especially if you’ve added documents or notes since the original generation

The first generation runs in the background after inspection and is ready in under 30 seconds. You can keep working on other things while it generates.

​

What’s in the report

The report has four sections:

​

1. Equipment summary

A line-by-line list of what went out and how it came back, including:

• Asset ID, item name, kit it belonged to (if any)
• Out date and Return date
• Final condition (Good / Damaged / Missing)
• Any condition notes you wrote at inspection

For kits, components are itemized. The kit overall shows whether all components returned cleanly.

​

2. Issues & damage

Any item marked Damaged or Missing during inspection:

• Item name and asset ID
• Condition assigned
• The notes you wrote
• Photos you attached (rendered inline in the report)
• Estimated repair / replacement cost (if you filled it in)
• Whether the cost was deducted from the deposit, charged separately, or absorbed

If there were no damage events, this section reads No issues recorded — clean return.

​

3. Customer notes & timeline

A digest of what happened during the job:

• Notes added by your team during On Site
• Any communications logged
• Any extensions or schedule changes
• Photos from delivery, on-site, or pickup

The timeline format makes it easy for someone reading the report later to reconstruct the job.

​

4. Recommendations

Claude AI suggests follow-up actions based on what happened:

• Inventory recommendations (“Camera Kit A’s UV filter has been damaged on 3 of the last 5 jobs — consider switching to higher-grade filters.”)
• Customer relationship suggestions (“Customer requested an extension twice in the last 6 months — they may be a candidate for a long-term contract.”)
• Operational follow-ups (“Outstanding deposit deduction of ฿1,200 needs to be reconciled in accounting.”)

These are AI-generated suggestions. They are useful starting points, not commands.

​

Accessing the report

​

Customer-facing version

The version visible in the customer portal is intentionally narrower:

• Equipment list with friendly names (no Asset IDs)
• Issues — only items they were told about and that affected their deposit (full transparency)
• A thank-you message
• A direct link to leave feedback

Internal cost estimates, recommendations, and any internal notes are NOT shown to the customer. This separation is enforced — there is no toggle or setting that exposes the internal sections to the portal.

​

Regenerating

To regenerate the report (after you’ve added photos, notes, or documents to the job): The previous version is preserved in the audit trail — you can compare versions if needed.

​

What the AI is good at

• Summarizing factual data into readable English / Thai prose
• Spotting patterns across closed jobs (e.g., a recurring damage type)
• Generating customer-friendly language from internal notes

​

What to verify yourself

• Cost figures. The AI uses the costs you typed at inspection. If the costs are wrong, the report is wrong.
• Customer-facing language. Read the customer-facing version before you assume the customer will understand it.
• Recommendations. They’re suggestions, not analyses backed by deep statistics. Use judgment.

​

Plan availability

The full AI Job Report is on the Operations plan and above. On Pro and Starter plans, jobs still close cleanly with a non-AI summary that lists equipment, durations, and damage flags but doesn’t generate the natural-language summary or recommendations.