intesol.kr
Cases / Manufacturing
Automation CASE 04

Report automation — 90% less writing time

A weekly production report that took a plant manager four hours became a 20-minute review. The numeric errors that came from hand-transcription dropped 91%.

90%
Less writing time
91%
Fewer numeric errors
4 weeks
Build time
Duration
4 weeks
Team
1 person
Started
2025 Q3
Market
Korea mid-size manufacturing
SITUATION

What was blocking them

ISSUE 01
Numbers were moved by hand
Figures pulled separately from MES, the quality sheet, and ERP were typed into Excel by the manager — over 200 cells a week, moved by a person.
ISSUE 02
Errors surfaced too late
Every quarter a transcription error turned up in a management meeting, and tracing the cause cost another half day.
ISSUE 03
The narrative changed every week
The same anomaly was described and weighted differently depending on who wrote it, which made trends hard to compare.
APPROACH

The order we worked in

Each phase began from what the previous phase measured. We kept that order because without the earlier step there is no way to verify the next one's effect.

01
Mapping data paths
We reconciled fields across MES, the quality sheet, and ERP to fix a single source per report line, resolving 14 duplicate definitions.
1 week
02
Aggregation pipeline
Aggregation and anomaly detection were automated, with supporting data attached to any line outside its threshold.
1.5 weeks
03
Narrative rules
Narrative paragraphs were fixed to a form extracted from 40 past reports. Nothing that is not grounded in a figure gets generated.
1 week
04
Handing over the review screen
We handed over a review screen the manager clears in 20 minutes, and ran two weeks in parallel.
0.5 weeks
RESULTS

Baseline at kickoff, measured after delivery

Metric
Before
After
Report writing time
4 h/week
→
20 min/week
Transcription errors
3.2 / week avg
→
0.3 / week avg
Anomaly detected at
Weekly meeting
→
Same day
Manually keyed cells
200+
→
0

How each figure is measured is set out in the measurement notes on the cases page. Revenue and cost figures come from the client's own accounting and are published only within the scope they approved.

“Writing the report turned into reading it. Knowing about anomalies before the meeting is the biggest change.”
Manufacturing · Production management lead
Report automation Claude API MES integration Excel integration Anomaly detection
Free · 1 business day
This one started with a diagnostic too
AI Workflow Automation starts by measuring your baseline. Find out where you stand first — the diagnostic is free.