EPS predictive policing / risk-terrain data analytics

Original source: https://thewalrus.ca/will-big-data-in-crime-fighting-create-a-new-era-of-racial-profiling/

Local snapshot (archived 2026-07-03): full-page PDF committed at sources/budgets/_archive/eps-predictive-policing-data-analytics/eps-predictive-policing-data-analytics.pdf.

• EPS chief analytics officer Stephane Contre and colleague Kris Andreychuk built an in-house predictive/risk-terrain system: compiled "233 distinct environmental and geographical features" on a digital map, overlaid with 2 years of 911-call + police crime data to find correlations.
• Risk terrain modeling (RTM): links environmental features (bars, abandoned buildings, transit stops, poor lighting, broken street furniture) to crime probability.
• Citizen Lab's "Algorithmic Policing in Canada" (2020) documents EPS among Canadian services exploring algorithmic/predictive tools.
• No Palantir or "ProvenTrue" use confirmed for EPS (Calgary Police uses Palantir Gotham; OPP also). EPS system appears largely in-house/internally developed.

Cost: NOT disclosed. Because it was internally developed by existing analytics staff, there is likely no large discrete procurement line — cost is embedded in salaries/analytics unit. Confidence: low on any dollar figure; treat as unpriced.