Leng: AI-Assisted Report Automation & Data Governance
Multi-department governance pipeline processing multi-owner submissions, applying local ML/AI attribution (llama.cpp/Qwen), and generating readiness scoring across 28 departments.
At Nile University, annual report generation and compliance tracking were previously fully manual processes spanning 28 distinct departments, creating severe bottlenecks, attribution ambiguities, and audit vulnerabilities.
Engineered an automated data governance pipeline and built Leng, an offline AI workbench (Electron + llama.cpp / Qwen) that routes submissions through conditional multi-step Laserfiche workflows and SQL Server ETL pipelines into executive Power BI reporting dashboards.
Integrated ML and AI-assisted source lineage and attribution verification, automatically auditing data provenance and generating departmental readiness scores with complete historical audit trails.
What was delivered
- Automated ETL pipeline from SQL Server into dynamic Power BI dashboards
- Laserfiche workflow automation with conditional logic and multi-step routing
- Offline AI workbench (Electron + llama.cpp/Qwen) for automated report synthesis & provenance auditing across 28 departments