Barrier-Free Housing · Wraparound Services · Unconditional Access
"Cloud by day. Fire by night. We travel with the people no other system will walk beside."
California spent roughly $24 billion on homelessness over five years, without consistently tracking whether the spending produced results. Communities, businesses, and emergency rooms carry the cost of unmanaged crisis every day. The people hardest to serve are the ones the current system is built to turn away.
Traditional shelter is built around requirements that screen out the people who need it most. The Prism & Pillar model removes those barriers — not as a loophole, but as the entire design.
To build unconditional residential campuses for the people the housing system has refused. Stability is honored as an outcome. Every resident is received in the condition they arrive in.
California's shelter system was designed for a narrow profile: temporarily displaced, sober, rule-compliant. The people who don't fit — those with active substance use, no ID, criminal histories, pets, mental health profiles incompatible with institutional intake — are excluded by design, not by accident.
The Prism & Pillar campus model is built on different assumptions entirely. Five named neighborhoods give residents a place and a community, and up to four unit types meet them where they are (tent, tiny home, vehicle, or beds as available). Two service tracks (Stability and Transition) operate independently of housing type. Biometric verification creates a stable identity that enables re-entry without restarting paperwork. No resident is ejected for behavioral incidents alone.
The Foundation is led by Dekari Smith-Russell, Acquisition Analyst at PSAI Realty Partners ($700M+ institutional real estate), law enforcement trained with senior SFPD relationships, and years of direct outreach in Bay Area encampments. This model is built by someone who has been paying close attention from both sides of every table.
The current system has one fatal flaw: a single disqualification sends a person back to the street with no alternative route. Every pathway in the Prism & Pillar model is built with parallel paths and feedback loops — so no one decision ends the journey.
Residents move up or step down between residency types without penalty, guided by a monthly review made with the resident, not for them. Re-entry is always open — a designed pathway with stored belongings, retained access, and an active case file behind it.
When behavior creates safety risk or disruption, the response is structured and proportional — never a single disqualification that sends someone back to the street. The goal is to keep people in the system, not to remove them. The track changes. The resident doesn't. Re-entry is a designed pathway, not an exception.
A person who steps down after a setback is not a failure of the system — it is the system working. The loop caught what the primary path could not sustain and rerouted them to where they can stabilize.
The population the Foundation serves carries real medical risk: overdose, cardiac events, falls. In a managed environment, a life-safety emergency is detected, located, and responded to in minutes, not discovered days later. Privacy is the default; it ends only when a life-sign drops.
A conceptual layout for the candidate Site 1 parcel on Washington Blvd, Fremont. Roughly 2.8 acres in the Irvington Transit Priority Development Area, organized around a single controlled entry. Permanent tiny homes sit on the Washington Blvd frontage, farthest from the active rail corridor; tents and vehicle pads occupy the rail side behind a planted buffer.
Candidate Site 1 · APN 525-342-9-5 · Washington Blvd, Fremont 94538 · ±2.8 acres · Irvington Transit Priority Development Area
Adjacent to the planned Irvington BART station expansion, currently stalled due to BART budget constraints · Active Union Pacific freight corridor on site boundary, sleeping units set back accordingly · Site control, ownership, and a City of Fremont zoning / use determination are pending
Photorealistic renderings to follow. This plan is the spec they build from.
Each residency type is its own neighborhood, with a name, a place on the map, and its own access protocols. Residents are received in the one that fits where they are, and move between them as needs change.
NEST is the family neighborhood — for families with minor children and for survivors fleeing harm. It activates in a later stage of Phase 1, once the core campus is stable. Families live here however they arrive: a tent, a vehicle, or a tiny home, with age-appropriate sleeping arrangements for the children inside each family's residency.
The campus is engineered for the people most systems route around. Targeted infrastructure is phased in across Phase 1 rather than activated all at once. The campus is not a clinical facility — serious mental-health conditions are supported through a harm-reduction model with referrals to outside clinical care.
Registered offenders subject to residency restrictions are accommodated at complementary Phase 2 sites, where family infrastructure is not present.
Intake is where the journey begins — and on this campus, intake is a security process. Every resident is met with the full human intake any shelter provides, plus a verified profile and biometric enrollment that anchors a stable identity from day one.
Once a resident is profiled at intake, they are matched the same day to one of up to four unit types, meeting them in the condition they arrive in. Residents move between types without penalty as stability changes.
Beds are added flexibly above this allocation as capacity allows.
Three phases with defined purposes and defined triggers — from a single Bay Area site to a proven model cities contract to deliver.
Establish the first site. Develop the operational model. Generate the first documented outcomes. Phase 1 scales across four stages from intake readiness to full capacity, on a single Bay Area site, opportunity-driven by whichever county opens the door first.
Grow the model across multiple Bay Area sites to roughly 335 residents, and refine it: making what worked in Phase 1 work better, and what didn't, work. Activate the hub-and-spoke architecture: centralized leadership, finance, development, technology, and donor functions serve every site, while site-specific roles scale with each new campus. This is why per-resident costs continue to fall.
Grow the network toward 545 residents. With the model proven, the Foundation partners with cities to provide the service each community needs, at the level they choose: implementing the managing model in an existing facility, developing a new campus site, or building out a regional network. Each engagement applies the same operational standard. The funding mix shifts to 60–70% government contracts as multi-year contracts mature.
The foundational two-lead structure: a CEO over an Operations Lead and a Residents Lead, each leading their teams. 20 employees + contracted security. Click the chart to zoom.
The model replicates: a central hub carries shared functions while each site runs the same two-lead unit. This is why cost per resident falls as the network grows. Click the chart to zoom.
The hub matures into a foundation. Centralized functions expand to support five sites and city partnerships, while each site continues to run the same two-lead unit. Click the chart to zoom.
Chronically unhoused individuals cost taxpayers $30,000–$50,000/year in public systems. We redirect those same dollars into a model that produces documented outcomes instead of unmanaged crisis.
Cost per resident is highest at opening and falls from $128 to $71 across Phase 1 as the campus reaches capacity — the model becomes more cost-effective the more people it serves. At Phase 3 scale across five sites, cost per resident reaches $62/day. Program service ratio: 90% at opening, 93% at full Phase 1 capacity. Food is budgeted at $7.50/resident/day as pure cash with no donations assumed; staffing includes 20% benefits; security is contracted; every stage carries a 2% operating reserve.
Stage 1 opening budget at 40 residents. Click any line to see how the number is built.
Site 1 candidate: a 2.8-acre Fremont parcel (Alameda County), built to 105 residents. Total Phase 1 capital is $3.15M, structured as four milestone-based drawdowns. The first drawdown ($2.2M) opens the campus at 40 residents. The remaining $996.5K is called across three subsequent drawdowns — $450K at 60 residents, $362K at 80 residents, $184.5K at full capacity — each triggered by demonstrated stability at the prior stage. Funds are never committed ahead of proof. Figures show the low–high range with midpoint.
Total cost per day to serve a given number of people. Doing nothing costs a flat $145 per person per day — the conservative Bay Area floor — so its total climbs steeply. Our total rises gently, because each added resident costs less than the last. The widening gap between the lines is the cost avoided. Move along the chart to read any point.
See how we reached $145–$217 per person per day — and adjust the assumptions yourself.
The funding mix evolves across phases — philanthropic-led in Phase 1, government-dominant by Phase 3 as multi-year contracts mature and outcomes are proven.
Both matter. The Foundation grows through two kinds of support: organizations that bring capabilities (Prism), and funders who bring capital (Pillar).
The Foundation is actively seeking founding partners, grant funding, government service contracts, and land access opportunities to launch Site 1 in the Bay Area.
Four pillars. One model. No dead ends.