How MobilityLens works
A step-by-step walkthrough of the individual journey, the case management lifecycle, and how research data flows — from first intake to population insight.
The individual journey
For anyone accessing MobilityLens directly
Set your location
Enter your ZIP code. Every recommendation, opportunity, and cost estimate is calibrated to your local market — not a national average.
Complete your assessment
Answer guided questions across all four capital areas: Human (skills, credentials), Social (network, community), Financial (income, savings, credit), and Resilience & Legal (crisis preparedness, legal situation).
Receive your EMS score
Your Economic Mobility Score (0–100) is calculated deterministically — no AI guesswork. You see exactly which factors drive your score and the estimated point gain from each action.
Build your goals
Choose which capital areas to strengthen first. Each goal breaks down into concrete steps with timelines, resources, and local providers — sorted by impact on your EMS.
Work with your AI advisor
Ask questions in plain language. Your advisor draws from Opportunity Insights, ALICE data, Social Capital Atlas, and local program databases — and always cites its sources.
Case management lifecycle
For nonprofits, social service agencies, and government case workers
Structured intake form captures demographics, household composition, income sources, presenting needs, and consent. Existing clients are matched automatically.
Caseworker completes a four-capital assessment. AI suggests areas of concern based on intake data. Risk flags surface automatically from case note language.
Worker creates a Case Plan with one or more Objectives. Each Objective has Goals, and each Goal has Milestones with due dates. The client sees their own goals in their individual dashboard.
Worker adds case notes, tracks referrals, monitors milestone completion, and receives alerts when the AI detects escalating risk language or stalled progress.
Case is closed with an outcome summary. De-identified outcome data joins the research pool — feeding aggregate insights for the next client in similar circumstances.
Case Plan hierarchy
Each milestone maps to a due date and a responsible party. The client sees their own milestones in plain language. The caseworker sees completion status and AI-generated priority alerts.
Research data flow
How closed cases become population-level insights — without exposing any individual
Case closed
A caseworker marks a case closed with an outcome summary (achieved, referred, withdrawn, or other). This triggers the de-identification pipeline.
De-identification
Name, contact details, case notes, and any identifying attributes are stripped. Only structured, categorical fields (age band, capital scores, outcome type, ZIP, household size bracket) enter the research pool.
Cohort check
Before any metric is surfaced in research views, the system counts distinct individuals in that cohort. If the count is below 50, the metric is suppressed — not approximated.
Research database
Approved records enter the research database, which is queryable by credentialed researchers, policymakers, and public dashboards. Zero PII. All fields are categorical bands, not raw values.
Benefits cliff detection
One of the most dangerous traps in economic mobility — and one of the hardest to see without data.
What is a benefits cliff?
A benefits cliff occurs when a raise in earned income causes a household to lose public benefits — SNAP, Medicaid, childcare subsidies, housing vouchers, or EITC — worth more than the income gained. The household ends up with less money than before the raise.
net_change = income_gained − total_benefits_lost
cliff = net_change < 0
How MobilityLens handles it
Every income recommendation is run through a cliff check before it reaches the user. If the check detects a cliff, the recommendation is blocked and replaced with a visual chart showing safe income targets — the exact dollar amounts where gains outpace losses.
Five benefits are modeled: SNAP, Medicaid, childcare subsidies, housing vouchers, and EITC — using the actual thresholds for the user's state and household size.
Frequently asked questions
- Is MobilityLens only for nonprofits?
- No. Individuals access MobilityLens directly at $5/month for the full AI advisor and four-capital plan. Nonprofits and government agencies use it for case management and population analytics. Researchers access a separate de-identified dataset view.
- What is the Economic Mobility Score (EMS)?
- The EMS is a 0–100 composite score across all four capitals plus a geographic modifier. It is calculated deterministically — not by an AI model. Think Experian credit score, but for economic mobility. Every point is traceable to a specific factor.
- What does "benefits cliff" mean?
- A benefits cliff happens when a raise in income causes a household to lose public benefits worth more than the raise itself — leaving them worse off financially. MobilityLens detects this automatically and blocks any income recommendation that would trigger a cliff, showing a visual chart of safe income targets instead.
- Can individuals share their data with their caseworker?
- Yes. Individuals can grant a specific caseworker read access to their EMS score and goals. Case note content and case plan details remain controlled by the NGO tenant. Nothing is shared without explicit consent.
- How is case data kept separate between organizations?
- Each organization operates in a fully isolated tenant. PostgreSQL Row Level Security enforces this at the database layer — not just in application code. No NGO can ever access another NGO's client records.
- What AI model powers the advisor?
- The AI advisor uses a local Mistral 7B model hosted on our infrastructure as the primary engine, with Claude API as an overflow fallback. No client data is sent to third-party APIs without explicit configuration from the organization.
- How do I know the AI recommendations are accurate?
- Every recommendation passes six validation checks: benefits cliff test, structural availability in your county, household suitability, ALICE data freshness (never older than 2 years), four-capital completeness, and citation presence. A recommendation without a citation is blocked.
- Who owns the client data inside an NGO's tenant?
- The NGO owns its client data in full. Xiente processes it only to run the platform. You can export or delete all data at any time. See our Privacy Policy and Terms of Service for the full data agreement.
- Is there a minimum cohort size for research data?
- Yes. No metric is displayed in research or analytics views unless the underlying cohort contains at least 50 distinct individuals. This prevents re-identification from small subpopulations.
- How does MobilityLens handle immigration-sensitive situations?
- Immigration status is never required to use the platform. Where relevant, the AI advisor surfaces immigration-aware pathways that do not require documentation disclosure. This data point, when voluntarily provided, is stored with heightened access restrictions.
See it live in 30 minutes.
We'll walk through the exact workflow for your organization type.
Request Demo