AI Leave Analysis
Reads an employee leave export, such as a CSV, and returns JSON with leave counts by type, month and department, the top 10 employees by leave days, and a written summary in English or Turkish.
Overview
AI Leave Analysis writes a report from employee leave records, such as an export from your HR system. The result shows how much leave was taken, of which kind and in which months, and, when the file has a status column, where each request stands in the approval flow.
The sample report on this page covers 155 leave records and 636 leave days. Its leave types are annual, sick, excused, maternity, marriage and paternity leave.
What you can build
- A monthly absence summary for HR, produced from the same export each time.
- An approval backlog check. The status section in the sample shows 13 requests waiting on a manager and another 13 marked pending.
- Planning notes for team leads on which leave types dominate and in which months and on which weekdays leave tends to start.
Inputs
Upload a leave export or paste its URL. Only the first file of a run is analyzed. A CSV export gets the full result with counts and tables. Other documents, such as PDF or DOCX, get only the written summary. A language setting, EN by default or TR, picks the language of the written report.
Outputs
One JSON file and a folder of CSV tables. The JSON opens with a written report in Markdown: an executive summary, key metrics, patterns over time, employee-level notes, leave types, and risks with recommendations. The computed figures follow, built from the columns the tool recognizes in your file: totals and averages, leave counts by type, department, month and weekday, status counts, and the top 10 employees by leave days and by number of requests. Each of these tables is also saved as a CSV. The sample on this page comes from an earlier version of the tool and has six sections:
- Summary statistics: total records and total leave days.
- Leave types, each with a count.
- Peak leave start dates, with how many leaves began on each.
- Top employees, three names with their totals.
- Status breakdown: approved, awaiting manager approval, pending and rejected.
- Insights: a few written observations, for example that annual leave makes up over half of all leave types.
Limitations
- In the sample, leave type and status labels stay in Turkish (Yıllık İzin, Onaylandı) while the insights are in English.
- The sample's per-type counts add up to 120 and its status counts to 128, against 155 total records. Reconcile the figures with your source file before using them for payroll.
- The top employees section lists people by name, so the report carries the same personal data as the file you upload.
API quick start
Run AI-Leave-Analysis with a single API call.
{
"inputDocumentMultiple": "https://your-cdn.com/input.png",
"language": "en"
}curl -X POST "https://api.wiro.ai/v1/Run/wiro/ai-leave-analysis" \
-H "Content-Type: application/json" \
-H "x-api-key: YOUR_WIRO_API_KEY" \
--data-binary @- <<'JSON'
{
"inputDocumentMultiple": "https://your-cdn.com/input.png",
"language": "en"
}
JSONOther models you might like
Explore similar wiro ai models.
