Operations overview
Maintenance posture, monthly budget allocation, and unresolved anomalies — April 2026.
Total machines
28
across 4 functional groups
Maintenance due soon
2
DUE_SOON · 7d window
Budget this month
R$242.092
8 cost centers · 202604
Flagged alerts
13
UNREVIEWED
Budget allocation by cost center
Planned spend · BRL · month 202604
Maintenance status
Distribution across 28 machines
28
Machines
Maintenance backlog
Machines past due or due within 7 days, sorted by urgency
| Machine | Type | Line | Days | Status |
|---|---|---|---|---|
| 3 | Producao | both | -70d | OVERDUE |
| 12 | Embalagem | both | -68d | OVERDUE |
| 14 | Producao | Line A | -68d | OVERDUE |
| 22 | Producao | both | -66d | OVERDUE |
| 26 | Producao | both | -64d | OVERDUE |
| 18 | Producao | Line B | -61d | OVERDUE |
| 27 | Producao | both | -60d | OVERDUE |
| 15 | Embalagem | both | -59d | OVERDUE |
Top anomalies
Pending review · sorted by deviation
Impostos · R$ 16.142,50
vs μ R$ 5.963,00
+71%
Fixas · R$ 5.365,00
vs μ R$ 2.009,00
+67%
RH · R$ 62.310,00
vs μ R$ 24.117,00
+58%
Fixas · R$ 4.977,50
vs μ R$ 2.009,00
+48%
Project
ADR · 2026-04
Anomaly detection is rule-based by design — zero labeled history makes ML inappropriate here.
Factory Data Platform
Portfolio project · Apr 2026
Built for a Brazilian food manufacturer operating 28 documented machines across two product lines. All controls previously ran on paper.
Stack
Kafka
Redis
BigQuery
dbt
Airflow
FastAPI
Docker
lakehouselakehouse.albuqr.com
monitortransaction-monitor.albuqr.com
apiplatform-api.albuqr.com
Data sources & known gaps
Documented, not fabricated
✓
Equipment
28 named and documented machines out of a larger fleet — remaining machines lack identification records.
✓
Budget
8 cost centers, real monthly allocations from DRE spreadsheet.
✗
Actuals
Source transactions had no cost center mapping — variance data unavailable. Documented, not fabricated.
✓
Alerts
Rule-based detection at 20% deviation threshold. Alerts shown are stored demo data; no live pipeline.