A geospatial intelligence platform designed to analyze and visualize broadband and fiber infrastructure across Nigeria.
# **udigap**
### *Geospatial Connectivity Intelligence Platform for Nigeria*
> **udigap** is a geospatial data engineering and analytics platform that maps, measures, and explains **fiber connectivity and broadband coverage** across Nigeria.
> It fuses **open spatial datasets**, **machine learning**, and **natural language querying** into one intelligent system — enabling planners, engineers, and policymakers to make data-driven infrastructure decisions.
---
## **Core Idea**
> Nigeria’s broadband expansion is accelerating, but spatial visibility into who is connected — and who’s left behind — remains fragmented.
> udigap bridges that gap by integrating datasets on **fiber routes**, **population**, **economic activity**, and **public facilities** to produce actionable connectivity insights.
You can ask udigap:
* “Which LGAs in Lagos have the lowest fiber-to-population ratio?”
* “Show me areas with high economic activity but no fiber coverage.”
* “What percentage of schools in Abuja are within 5 km of fiber infrastructure?”
* “Where should we prioritize fiber deployment for maximum economic impact?”
---
## **System Architecture**
```
External Datasets (NCC • OSM • GRID3 • VIIRS)
│
▼
ETL Pipeline (Python / AWS Lambda)
│
▼
S3 Data Lake ───────────────► PostgreSQL + PostGIS
│ │
▼ ▼
Optional ML Layer ───────────────► FastAPI + LangChain
│ │
▼ ▼
Predictions (fiber_priority_zones) → Streamlit Dashboard
```
---
## **Repository Structure**
```
udigap/
├── data/ # raw / processed / analytics / parquet layers
├── etl/ # extraction, transformation, and loading scripts
├── ml/ # optional machine learning layer
├── api/ # FastAPI + LangChain NL-to-SQL service
├── dashboard/ # Streamlit web interface
├── sql/ # database schema, indexes, and materialized views
├── docker/ # contain …