Case Study: HarrisLogic achieves 99% accurate recidivism predictions and reduced crisis spending with SAP HANA

HarrisLogic Protecting Our Communities with Predictive Data

HarrisLogic, a St. Louis–based high‑tech company with 220 employees that builds software for healthcare clinical operations and systems integration, set out to help governments predict whether people with mental‑health conditions who had been jailed would return to custody within six months. The challenge was to integrate seven disparate data sources — from mental‑health records to criminal‑justice data — and deliver an engaging, actionable user interface for government and clinical users.

Using SAP Predictive Analytics and the SAP HANA platform on AWS (with partner EV Technologies), HarrisLogic deployed real‑time predictive models (native logistic regression) that flag key risk factors for clinicians and deliver up‑to‑the‑minute predictions on incoming jail records. The solution processes about 12,000 encounters per month, achieves 99% prediction accuracy, helped cut behavioral‑health crisis spending by 25%, and supports higher crisis‑diversion rates (about 89% in major metros), with plans to expand into new markets.

HarrisLogic

Hudson Harris

Chief Privacy Officer