The Problem

In East Africa, communities often face a lack of reliable, accessible tools for documenting and reporting incidents. Individual incidents are often isolated and go unreported, making it difficult to identify larger patterns of injustice, protect human rights, and organize collective action.

The Solution

I developed "Watchtower," a multitenant, AI-powered incident reporting platform that enables individuals to securely document events and aggregates these isolated reports into collective community knowledge. The platform uses AI to analyze patterns and anomalies in the data, transforming raw reports into actionable intelligence. This processed information is published on interactive maps, empowering communities to take informed action.

My Role & Responsibilities

As a full-stack developer I built the data ingestion pipeline, which securely processes incident reports from a variety of sources. I built the back-end logic for the AI-powered contextual analysis, working to detect patterns and anomalies. On the front-end, I implemented the interactive maps data visualizations, and comprehensive dashboards for all user groups.

The Technology Stack

Front-End: React, Mapbox, Tailwind

Back-End: TRPC, Typescript, node

Database: PostgreSQL

Mapping API: Mapbox,

The Outcome/Impact

Watchtower provides a critical tool for civic participation and safety in East Africa. By aggregating and analyzing incident data, it helps communities and human rights organizations identify trends, allocate resources effectively, and advocate for change. The platform's success is measured by the number of incidents reported and the tangible community actions that have resulted from the data-driven insights it provides.