DOOR 05·KREVIA SPECIALTY

AI & GIS Systems

DATA THAT EXPLAINS ITSELF

Geospatial intelligence with practical deep learning — from municipal waste WebGIS to AI invoice parsing and inventory forecasting.

Deliverables & Methodology
Data modelling & spatial pipelines
Map-based dashboards & Leaflet/Mapbox
AI-assisted automation
Smart parsing & validation
Deployment & documentation
FILTER BY TECH:
ARCHITECTURE
Pekanbaru GIS Spatial Zoning EngineKREVIA SYSTEM ARCHIVE
2025AI & GIS

Pekanbaru GIS Spatial Zoning Engine

Urban spatial modeling engine rendering multi-polygon RTRW zoning and demographic waste density maps for Pekanbaru.

⚡ Sub-second GeoJSON polygon layer rendering
OpenLayersSpatial ModelingNext.js 14+1
ARCHITECTURE
AI Invoice Maker (Smart Local Parser)KREVIA SYSTEM ARCHIVE
2025AI & GIS

AI Invoice Maker (Smart Local Parser)

Smart Local Parser engine that uses pattern matching and heuristic tokenization to extract invoice items from unstructured text.

⚡ Zero-cloud local processing & 28% time savings
Heuristic NLPToken ExtractionClient-side AI+1
ARCHITECTURE
StockSense AI (Demand Forecasting)KREVIA SYSTEM ARCHIVE
2025AI & GIS

StockSense AI (Demand Forecasting)

AI decision intelligence platform for restaurants to optimize ingredient purchasing, eliminate stockouts, and minimize food waste.

⚡ MAPE < 15% Forecast Accuracy
Time Series ForecastingPython FastAPIMachine Learning+1
ARCHITECTURE
ProjectDL (FashionAI GRU Recommendation)KREVIA SYSTEM ARCHIVE
2024AI & GIS

ProjectDL (FashionAI GRU Recommendation)

Sequential deep learning model using Gated Recurrent Units (GRU) to predict next-item fashion purchases from session history.

⚡ Low-latency sequential session inference
PyTorchGRU Neural NetworkFastAPI+1