AI Systems Online
AI Command Center
Real-time monitoring of the AI/ML pipeline powering the Lao Community Data Portal. 4 predictive models, 3 NLP engines, and 6 data sources working together to transform reactive crisis response into proactive, culturally-grounded community support.
0
Active ML Models
0
Predictions Generated
0%
Avg Model Accuracy
1
Fairness Reviews
AI Pipeline Architecture
LIVEData Ingestion
6 sources
NLP Pipeline
3 engines
ML Models
4 predictive
Fairness Audit
Continuous
Early Warning
Real-time
Intervention
Recommended
Predictive Models
Live Activity
Pipeline Throughput
NLP Engines
Bilingual Sentiment Analyzer
LLM-Enhanced (Claude + Lexicon)
Community Topic Extractor
LLM-Enhanced Topic Modeling
PII Anonymization Pipeline
NER + Rule-Based
Neural Network Visualization
ACTIVEMLP Architecture — Mental Health Crisis Predictor
Model Performance Overview
Data Pipeline Status
| Data Source | Records | Last Sync | Frequency | Status |
|---|---|---|---|---|
| Census ACS 5-Year | 15,247 | 2026-04-12 | Annual | Synced |
| MDH BRFSS | 3,842 | 2026-04-10 | Quarterly | Synced |
| Community Survey 2025 | 1,205 | 2026-03-01 | Annual | Synced |
| Community Narratives | 2,847 | 2026-04-13 | Continuous | Live |
| MN Dept of Education | 8,920 | 2026-04-08 | Semester | Synced |
| BLS Employment Data | 4,310 | 2026-04-11 | Monthly | Synced |
Ethical AI Framework
Human-in-the-Loop
All predictions require human review before intervention. No automated actions on vulnerable populations.
Fairness Monitoring
Continuous disparate impact testing across gender, age, and generation. Four-fifths rule enforced on all models.
Community Governance
Community IRB reviews all model deployments. Data sovereignty maintained through consent framework.
AI/ML Pipeline Monitoring System · Lao Community Data Portal · Last updated: 4/15/2026