About BlightLens
An AI-powered property risk platform that helps the City of Tulsa and Housing Forward find at-risk properties before they become a community burden. After a 2025 pilot in North Tulsa, the 2026 cohort is scaling to capture and analyze every parcel in the city.
What We’re Building
BlightLens combines parcel data, code-enforcement records, county tax and ownership records, and street-level imagery captured by our students into a single property risk profile. For every parcel in Tulsa we compute a deterministic composite score from publicly available signals — open demolition cases, unresolved maintenance violations, nuisance complaints, major repair needs, and tax liens — and surface the properties most likely to need intervention.
The platform is used by analysts at Housing Forward and the City of Tulsa to triage cases, allocate resources, and document the why behind every action. Every score is traceable to the specific records that produced it.
From a North Tulsa Pilot to Citywide
In summer 2025 we ran the first version of BlightLens against the InvestNorth area of North Tulsa — roughly 6,300 parcels in an InvestNorth-funded geography. That focus let us build the data contracts, the risk model, and the partnerships with Housing Forward and the City of Tulsa we needed to do this responsibly.
In summer 2026, the cohort is scaling to the entire city. Students drive every street and capture 360° imagery of every parcel, and the risk model runs against the full Tulsa parcel set. The 60%-by-2028 commitment applies citywide — and so do we.
Tulsa Is the Laboratory
We’re not building a Tulsa-only tool. We’re building a methodology that any city can adopt — but Tulsa is where it gets tested at scale, with real partners, against a public commitment we’re measured on. The data contracts, the risk model, the imagery pipeline, and the community-review workflow are designed from day one to be portable.
When the methodology graduates from this lab, the next city should be able to stand up their own BlightLens with their own parcel data and their own civic partners — not start from scratch.
The Partnership
Housing Forward
Primary community partner. Brings local expertise, the operational workflow, and direct relationships with Tulsa residents.
City of Tulsa
Development Services and code enforcement, ensuring the platform aligns with municipal housing and development goals.
University of Tulsa
Center for Real Estate Studies and the TURC student program. Research direction, data engineering, and the bridge to community partners.
Harvard GSD
Faculty in real estate and urban economics anchor the academic methodology and ensure the work contributes to the broader housing-policy conversation.
Research and Student Work
BlightLens is the foundation for forthcoming research in the Journal of Housing Economics special issue on local public finance and housing. The dataset, risk model, and review workflow are also the curriculum for the TURC (Tulsa Undergraduate Research Challenge) student cohort — undergraduate and graduate researchers who work directly with community partners on real properties, real decisions, and real impact.
Students rotate across 3 teams: property analysis (intelligence and feasibility), resource matching (incentives and development resources), and project coordination (community connection and integration). The arrangement gives students a portfolio of applied work and gives the project the throughput it needs to keep up with the city.
Operating Principles
Community-centered
Built with Housing Forward, the City of Tulsa, and Tulsa residents. The model serves the people who already do this work, not the other way around.
Transparent by design
Risk scores are deterministic and auditable: every input is a public record we can show on the property page. No black-box predictions.
Academic rigor
Methodology and findings contribute to the Journal of Housing Economics special issue on local public finance and housing, and are open to peer review.
Replicable
Tulsa is our laboratory. The data contracts, risk model, and review workflow are designed so other cities can adopt the same approach once we've proven it here.
Want to Know Who’s Behind This?
The faculty, students, and partners actively building BlightLens are listed on the team page.