I’m Alfred Kinoti, a researcher at Metahac Labs focused on building data-driven intelligence systems for public safety. My work sits at the intersection of criminology, data science, and field observation — translating raw incident data into predictive tools that law enforcement can actually use.
The VMIP (Violence and Misconduct Intelligence Platform) framework is my primary research vehicle. In 2025, I compiled and analysed 500 crime incident records across 17 Nairobi sub-counties, building a GBM sigmoid model that generates per-zone P(intent) scores with an AUC of 0.891.
Beyond the data, I conduct structured field observations in high-density areas — Dandora, Eastleigh, Mathare, Westlands — to ensure the models reflect ground-level realities that datasets alone can miss.
