Elder Care Facility Intelligence MCP Server

Elder care facility intelligence delivered as a Model Context Protocol server — assess any nursing home, assisted living, or memory care facility in seconds from your AI client. This MCP server orchestrates 9 parallel data sources to produce composite safety scores, complaint pattern analysis, ownership transparency ratings, quality composites, and staff adequacy estimates. Purpose-built for private equity due diligence, insurance underwriting, family placement decisions, and regulatory oversight.

The server runs in Apify Standby mode, staying active between requests for sub-second tool invocation. Each call dispatches parallel requests to OSHA inspections, CFPB consumer complaints, multi-platform review scrapers, corporate registries, nonprofit databases, and Google Maps — then synthesizes raw data into structured JSON reports with 0-100 scores, categorical verdicts, and actionable signals. No API keys to manage, no infrastructure to maintain.

What data can you extract?

Why use Elder Care Facility Intelligence MCP Server?

Manual facility due diligence means pulling OSHA records from one portal, CFPB data from another, searching OpenCorporates for ownership chains, reading through dozens of reviews on three separate platforms, and synthesizing it all by hand. For a single facility, that process takes 4-6 hours. For a portfolio of 20 facilities, it takes weeks.

This MCP server automates the entire process. Call one tool, get a structured report with scores, signals, and verdicts in under 30 seconds. Every data source runs in parallel — no sequential API polling, no copy-pasting between browser tabs.

Beyond the speed advantage, the server provides benefits that come from running on the Apify platform:

• Scheduling — run recurring facility monitoring on daily, weekly, or custom intervals to track deteriorating safety metrics over time
• API access — trigger assessments from Python, JavaScript, or any HTTP client for integration into existing workflows
• Standby mode — the server stays warm between requests, eliminating cold-start latency for high-frequency use cases
• Monitoring — get Slack or email alerts when runs fail or produce unexpected results
• Integrations — connect assessment outputs to Zapier, Make, Google Sheets, HubSpot, or downstream webhooks

Features

• 4 scoring models + composite verdict — Facility Safety Score, Complaint Severity Index, Ownership Transparency Rating, and Quality Rating Composite combine via weighted formula into a single 0-100 composite risk score with RECOMMENDED / ACCEPTABLE / CAUTION / AVOID / HIGH_RISK verdict
• OSHA violation classification — differentiates serious, willful, and repeat violations with separate penalty accumulators; willful violations carry 3x the weight of serious violations in the safety score (15 pts vs 5 pts per event)
• Violation density analysis — calculates violations-per-inspection ratio to identify facilities with high citation rates relative to inspection frequency, not just raw counts
• CFPB complaint severity triage — classifies each complaint as high/medium/low severity based on company response status ("closed without relief" = high, "closed with explanation" = medium); filters for elder care billing keywords
• Cross-platform complaint correlation — detects when high-severity CFPB complaints coincide with negative review patterns and flags systemic issues
• Staffing keyword NLP — scans review and complaint text for 10 staffing-related terms (understaffed, short-staffed, neglect, abandon, response time, etc.); 5+ mentions triggers understaffing signal, 10+ triggers critical concern
• Nonprofit status verification — cross-references ProPublica IRS 990 data to confirm tax-exempt status and report revenue size; for-profit operators receive an additional risk flag
• Corporate complexity scoring — counts jurisdictions, dissolved entities, holding/management company structures, and shell indicators; assigns complexity points per entity type (jurisdiction +3 pts, dissolved +5 pts, shell indicator +4 pts)
• Review consistency scoring — compares average ratings across Google Maps, Trustpilot, and multi-platform aggregators; cross-platform divergence above 1.0 stars triggers inconsistency signal
• 8 specialized MCP tools — from a focused regulatory history query to a full 9-source composite assessment, granular tool selection lets callers pay only for the intelligence they need
• Parallel actor dispatch — all data sources run concurrently via Promise.all, reducing a 9-source composite assessment to the latency of the slowest single source (~15-30 seconds)
• Spending limit enforcement — every tool call checks Actor.charge() before executing and returns a structured error if the per-run budget ceiling is reached

Use cases for elder care facility intelligence

Private equity acquisition due diligence

PE firms evaluating nursing home or assisted living acquisitions need to surface compliance risk before closing. This MCP server provides the OSHA violation history, complaint trajectory, and ownership structure analysis that may not appear in seller data rooms. Call facility_safety_assessment for target facilities in parallel to rank acquisition candidates by composite risk score, then use ownership_structure_check to identify layered holding company structures before negotiating representations and warranties.

Insurance risk underwriting

Underwriters pricing liability and workers' compensation policies for elder care operators need objective safety and staffing indicators beyond loss run history. The regulatory_violation_history tool surfaces willful and repeat OSHA violations — the strongest predictors of ongoing liability exposure. The staff_adequacy_estimate tool provides a staffing adequacy proxy from public signals, without requiring access to internal HR records.

Family placement decisions

Families researching nursing homes or assisted living communities for a parent or spouse face information asymmetry — facility marketing materials do not disclose OSHA violations or unresolved billing complaints. The quality_rating_composite tool aggregates ratings from Google Maps, Trustpilot, and other review platforms into a single score, and complaint_pattern_analysis surfaces billing and care complaints from the CFPB database that most families never think to check.

Portfolio risk monitoring

Operators and investors managing multiple elder care facilities can call compare_facilities for each property on a schedule to track safety scores, complaint volumes, and quality ratings over time. Rising complaint severity or a new OSHA willful violation triggers an early warning before it becomes a regulatory action or adverse media event.

Regulatory compliance oversight

State regulators and advocacy organizations monitoring elder care quality can use regional_facility_scan to discover all facilities in a geographic area, then run facility_safety_assessment on facilities with low Google Maps ratings to triage inspection resources toward the highest-risk operators.

Staffing adequacy auditing

Advocacy groups and labor organizations tracking understaffing in elder care can use staff_adequacy_estimate to extract staffing signals from public review and complaint data at scale. The tool cross-references review keyword frequency with OSHA staffing violation counts to identify facilities where understaffing indicators appear across multiple independent data sources simultaneously.

How to assess an elder care facility

1. Connect the MCP server to your AI client — Add the server URL https://elder-care-facility-intelligence-mcp.apify.actor/mcp to Claude Desktop, Cursor, Windsurf, or any MCP-compatible client. Your Apify API token must be included as a Bearer header or query parameter.
2. Choose your tool — For a full picture, use facility_safety_assessment. For a focused query, choose from seven specialized tools. The tool list appears automatically in your MCP client after connecting.
3. Provide facility name and location — Type the facility name (e.g., "Sunrise Senior Living") and an optional city/state for geographic disambiguation. The server constructs the appropriate search queries for each data source automatically.
4. Review the structured report — Results return as JSON with 0-100 scores, categorical risk levels, natural-language signals, and recommendations. A full composite assessment typically completes in 15-30 seconds.

MCP tools

Connection configuration

Claude Desktop — add to claude_desktop_config.json:

{
  "mcpServers": {
    "elder-care-facility-intelligence": {
      "url": "https://elder-care-facility-intelligence-mcp.apify.actor/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_APIFY_TOKEN"
      }
    }
  }
}

Cursor / Windsurf / Cline — add the same URL and Authorization header in your MCP settings panel.

HTTP / programmatic — POST directly to the endpoint:

curl -X POST "https://elder-care-facility-intelligence-mcp.apify.actor/mcp" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_APIFY_TOKEN" \
  -d '{"jsonrpc":"2.0","method":"tools/call","params":{"name":"facility_safety_assessment","arguments":{"facility":"Sunrise Senior Living","location":"Virginia"}},"id":1}'

Output example

A facility_safety_assessment call for "Harborview Care Center, Tampa FL" returns:

{
  "facility": "Harborview Care Center",
  "compositeScore": 63,
  "verdict": "AVOID",
  "facilitySafety": {
    "score": 72,
    "oshaViolations": 7,
    "seriousViolations": 4,
    "repeatViolations": 2,
    "riskLevel": "POOR",
    "signals": [
      "2 repeat violations — pattern of non-compliance",
      "4 serious violations — significant safety concerns",
      "$31,500 in OSHA penalties"
    ]
  },
  "complaintSeverity": {
    "score": 55,
    "totalComplaints": 18,
    "severityBreakdown": { "high": 4, "medium": 6, "low": 8 },
    "reputationLevel": "POOR",
    "signals": [
      "4 unresolved/untimely CFPB complaints — consumer protection concern",
      "Average rating 2.8/5 — poor resident/family satisfaction",
      "11 negative reviews — reputation concern",
      "Correlated complaints across CFPB + review platforms — systemic issue"
    ]
  },
  "ownershipTransparency": {
    "score": 48,
    "entitiesFound": 5,
    "nonprofitStatus": false,
    "corporateComplexity": 31,
    "transparencyLevel": "OPAQUE",
    "signals": [
      "3 jurisdictions — complex ownership structure",
      "2 dissolved entities — ownership history concern",
      "3 holding/management entities — potential layered ownership"
    ]
  },
  "qualityRating": {
    "score": 34,
    "avgRating": 2.8,
    "reviewCount": 63,
    "hasPhysicalPresence": true,
    "qualityLevel": "BELOW_AVERAGE",
    "signals": [
      "Google Maps 2.9/5 — poor local reputation"
    ]
  },
  "allSignals": [
    "2 repeat violations — pattern of non-compliance",
    "4 serious violations — significant safety concerns",
    "$31,500 in OSHA penalties",
    "4 unresolved/untimely CFPB complaints — consumer protection concern",
    "Average rating 2.8/5 — poor resident/family satisfaction",
    "Correlated complaints across CFPB + review platforms — systemic issue",
    "3 jurisdictions — complex ownership structure",
    "2 dissolved entities — ownership history concern",
    "Google Maps 2.9/5 — poor local reputation"
  ],
  "recommendations": [
    "Severe complaint pattern — review resident rights and grievance procedures",
    "Opaque ownership — request beneficial ownership disclosure",
    "Repeat violations indicate systemic safety management failure",
    "For-profit facility — verify pricing transparency and care-to-revenue ratios"
  ]
}

Output fields

How much does it cost to assess elder care facilities?

Elder Care Facility Intelligence MCP Server uses pay-per-event pricing — you pay $0.045 per tool call. Platform compute costs are included. There is no subscription, no minimum spend, and no charge for idle standby time.

You can set a maximum spending limit per run to control costs. The server stops processing when your budget is reached and returns a structured error message rather than a partial result.

For comparison, manual OSHA and CFPB research services charge $50-200 per facility report. CMS-based nursing home data platforms with similar risk scoring charge $300-800/month for subscription access.

How Elder Care Facility Intelligence MCP Server works

Data collection phase

When a tool is called, runActorsParallel() dispatches concurrent Actor.call() requests to between 1 and 9 downstream Apify actors depending on which tool was invoked. Each actor call is allocated 512 MB memory and a 120-second timeout. All calls run via Promise.all(), meaning the response time equals the slowest single actor, not the sum of all actors. Raw results are collected as arrays of items from each actor's default dataset.

Scoring phase

Raw data flows through four independent scoring functions in scoring.ts. Each function is designed around elder care risk factors:

scoreFacilitySafety parses OSHA inspection records and classifies each violation by type string matching ("serious", "willful", "repeat"). The penalty formula weights willful violations at 15 points each, repeat violations at 10 points, and serious violations at 5 points, capped at 50 from violation counts. An additional violation density sub-score (max 25) normalizes citation counts against the number of distinct inspection events using a Set deduplication on inspection IDs. A recency sub-score (max 10) adds 3 points for violations within the past 12 months and 1 point for violations within 24 months.

scoreComplaintSeverity processes CFPB complaint records and classifies them by company response string: "closed without" or "untimely" = high severity, "closed with explanation" = medium. It filters complaints by 8 elder care billing keywords. Review data from multi-platform and Trustpilot sources contributes a review risk sub-score calculated as (5 - avgRating) * 7. A cross-platform correlation check fires a "systemic issue" signal when high CFPB severity and negative review volume both exceed thresholds simultaneously.

scoreOwnershipTransparency builds a complexity score by iterating all corporate entities and accumulating points: 3 per jurisdiction, 5 per dissolved entity, 4 per holding/management/trust/LLP entity type, and 20 if no entities are found at all. Nonprofit status confirmed via ProPublica data reduces risk (score contribution = 0 rather than 15). Web presence gaps from Website Contact Scraper and Google Maps add up to 25 additional opacity points.

scoreQualityRating is the only inverted metric — higher score means better quality. It aggregates ratings across review sources, weighting the average by 8 (max 40), adding Google Maps rating points (max 25), factoring review volume (max 20), and adding up to 15 points for cross-platform rating consistency when the divergence between average review rating and Google Maps rating is under 0.5 stars.

Composite assembly

generateElderCareIntel() applies weighted combination: safety (30%), complaint severity (25%), ownership opacity (20%), and inverted quality rating (25%). This weighting reflects that safety violations carry the highest regulatory and liability consequence, followed by complaint patterns, then ownership risk, with quality serving as a positive counter-signal. Signals from all four models are merged into allSignals and a set of conditional recommendations is generated based on categorical threshold crossings in each scoring dimension.

Tips for best results

1. Include city and state for common facility names. Chains like "Sunrise Senior Living" or "Brookdale" operate hundreds of locations. Adding "Phoenix AZ" or "Tampa FL" as the location parameter scopes OSHA and Google Maps queries to the specific site.

2. Use regulatory_violation_history for OSHA-only queries. If you only need violation data, this tool is faster and cheaper than the full composite. Call the composite facility_safety_assessment only when you need all four scoring dimensions.

3. Run regional_facility_scan first for geographic discovery. When you do not have a shortlist of facility names, scan the region first to get a ranked list of local facilities with ratings, then run targeted assessments on the lowest-rated results.

4. Cross-reference ownership entity names. The ownership_structure_check tool returns corporate entity names from OpenCorporates. Running a second assessment with the parent company name (rather than the facility name) often surfaces additional corporate structure data.

5. Batch calls across facilities for portfolio screening. If you are evaluating 10 facilities simultaneously, call compare_facilities for each in parallel from your orchestration layer rather than sequentially — total wall time stays the same as a single assessment.

6. Set a per-run spending limit. Use Apify's run-level budget controls when calling from automated workflows to prevent unexpected spend from loops or retries.

7. Check nonprofit status first for pricing transparency investigations. If the nonprofitStatus field is false, the operator is for-profit and the revenue-to-care ratio concern in recommendations is actionable — ProPublica 990 data will be absent and ownership chain scrutiny is warranted.

Combine with other Apify actors and MCP servers

Limitations

• No CMS Nursing Home Compare integration. The Centers for Medicare and Medicaid Services star rating system is not directly queried. The quality_rating_composite provides a comparable consumer-signal rating but does not replicate CMS five-star methodology.
• Facility name disambiguation is imperfect. National chains operating hundreds of locations under the same brand name may return aggregated data across multiple properties. Always provide city and state to narrow results.
• OSHA data covers workers, not residents. OSHA inspections address workplace safety for employees. Resident-specific regulatory data (state health department surveys, deficiency citations) is not included — these databases do not have publicly accessible APIs.
• CFPB complaints reflect financial services, not care quality. The CFPB dataset captures billing, debt collection, and financial product complaints. Clinical care complaints filed with state health departments are outside this MCP's scope.
• Review data reflects self-selection bias. Families who have strongly positive or negative experiences are overrepresented in public reviews. Low review volume (under 10 reviews) reduces scoring reliability.
• Corporate registry coverage varies by jurisdiction. OpenCorporates covers 140+ jurisdictions but depth varies. Small regional operators in less-documented jurisdictions may show fewer entity matches than their actual corporate structure warrants.
• Nonprofit data relies on IRS 990 filing recency. ProPublica IRS 990 data may lag 12-18 months behind the current fiscal year. Revenue figures and executive compensation are historical.
• No real-time staffing ratios. Staff adequacy estimation is a proxy derived from review keyword NLP and OSHA staffing violations — it does not access payroll records, shift logs, or state staffing mandate compliance reports.

Integrations

• Apify API — trigger facility assessments programmatically from Python, JavaScript, or any HTTP client; parse JSON output directly into your risk management workflows
• Webhooks — receive notifications when scheduled facility monitoring runs complete or return high-risk verdicts for immediate downstream action
• Zapier — route AVOID or HIGH_RISK verdicts to Slack channels, email alerts, or CRM deal stage updates automatically
• Make — build multi-step workflows combining facility assessment output with document generation, calendar scheduling, or case management systems
• Google Sheets — pipe facility scores and signals into a live monitoring dashboard for portfolio tracking or comparative analysis
• LangChain / LlamaIndex — use facility assessment JSON as context for RAG pipelines, investment memo generation, or AI-assisted due diligence report writing

Troubleshooting

• Returns empty OSHA data for a facility. OSHA records are indexed by establishment name and state. If the operating entity name differs from the facility's public marketing name, try the legal operator name (often visible in the ownership_structure_check corporate results) as input to regulatory_violation_history.

• Composite score seems low despite known problems. The quality rating dimension is inverted — facilities with good review scores partially offset high safety and complaint risk scores in the composite. Review the individual dimension scores and signals in the response rather than relying solely on compositeScore when signals in a single dimension are the primary concern.

• Ownership structure returns zero entities. Some smaller regional operators use DBAs or trade names that differ from their registered legal entity names. Re-run ownership_structure_check with the legal entity name from state licensing records or from the facility's own website footer for better corporate registry match rates.

• regional_facility_scan returns fewer facilities than expected. The scan uses Google Maps and returns up to 20 results per query. Specify the facilityType parameter ("nursing home", "assisted living", or "memory care") to narrow results to a specific care category and reduce mixed-type results.

• Spending limit reached error. The tool returns {"error": true, "message": "Spending limit reached for [tool_name]"} when the per-run budget ceiling is hit. Increase your Apify run-level spending limit in the Apify Console or reduce the number of tool calls per session.

Responsible use

• This MCP server only accesses publicly available government databases and consumer review platforms.
• OSHA and CFPB data are published by US federal agencies for public access and accountability purposes.
• Do not use facility assessment outputs as the sole basis for regulatory action, legal filings, or public accusations without independent verification.
• Scores and signals are derived from automated analysis of public records and may not reflect the current state of a facility's operations.
• For guidance on web scraping legality, see Apify's guide.

FAQ

How many elder care facilities can I assess in one session? There is no per-session limit. Each tool call is billed at $0.045. You can call tools in parallel from your orchestration layer to screen dozens of facilities simultaneously. Apify's free plan includes $5/month in credits, covering approximately 110 facility assessments.

Does elder care facility intelligence include CMS star ratings? No. CMS Nursing Home Compare star ratings are not directly integrated because CMS does not provide a public API for that data. The quality_rating_composite tool provides a comparable consumer-signal quality rating from Google Maps, Trustpilot, and multi-platform review data, but the methodology differs from CMS five-star scoring.

How accurate is the OSHA safety score for nursing homes? The safety score reflects violations in the OSHA inspection database, which covers workplace safety for facility employees. Accuracy depends on OSHA database completeness for the establishment name provided. Providing the legal entity name and state improves match rates. The score is an indicator, not a regulatory determination.

How is this different from CMS Nursing Home Compare or Medicaid/Medicare databases? CMS databases require web scraping or licensed data agreements and focus on clinical care deficiencies. This MCP combines OSHA (worker safety), CFPB (financial complaints), public corporate registries (ownership structure), and consumer reviews — data dimensions that CMS does not cover and that are particularly relevant for financial and operational due diligence.

Can I use elder care facility intelligence for assisted living and memory care, not just nursing homes? Yes. All tools accept any facility name. The regional_facility_scan tool has an explicit facilityType parameter that accepts "assisted living", "memory care", "nursing home", or any other care type string. Assisted living facilities are subject to OSHA and may have CFPB complaints from billing disputes.

How long does a full facility safety assessment take? The facility_safety_assessment tool dispatches 9 parallel actor calls. Total response time is typically 15-30 seconds, equal to the slowest single data source. The server runs in Apify Standby mode so there is no cold-start delay between calls.

Is it legal to use OSHA and CFPB data for facility research? Yes. OSHA inspection data and CFPB consumer complaints are published by US federal agencies specifically for public accountability. Using publicly available government databases for research and due diligence is legal. See Apify's guide on web scraping legality for broader context.

Can I detect understaffing before it becomes an OSHA violation? The staff_adequacy_estimate tool provides a leading-indicator proxy by extracting staffing-related keywords from consumer reviews and CFPB complaints — data sources that often reflect staffing problems before OSHA receives a formal complaint. Five or more staffing mentions across reviews and complaints triggers an understaffing signal; ten or more triggers a critical concern flag.

What MCP clients does this server work with? The server implements the Model Context Protocol over HTTP with StreamableHTTP transport. It works with Claude Desktop, Cursor, Windsurf, Cline, and any other MCP-compatible client that supports HTTP-based MCP servers. Programmatic access is also available via direct HTTP POST requests.

Can I schedule recurring facility monitoring on a portfolio? Yes. Use the Apify scheduling feature to trigger tool calls on a daily, weekly, or custom interval. Combine with webhooks to receive alerts when composite scores cross risk thresholds, enabling proactive monitoring without manual re-checking.

What happens if a facility has no OSHA records? If no OSHA records are found, the safety dimension scores near zero (low risk). This may reflect a genuinely clean compliance record or the result of a name mismatch. Check the ownershipTransparency output to confirm you have the correct legal entity name, then retry with the registered operator name if needed.

How is ownership transparency scored for for-profit versus nonprofit operators? Nonprofit operators confirmed via ProPublica IRS 990 data receive 0 points on the nonprofit sub-score (lower risk). For-profit operators receive 15 additional risk points. Nonprofit status is not treated as a risk elimination — subsequent corporate complexity scoring still penalizes layered ownership structures even in nonprofit chains.

Help us improve

If you encounter issues, you can help us debug faster by enabling run sharing in your Apify account:

1. Go to Account Settings > Privacy
2. Enable Share runs with public Actor creators

This lets us see your run details when something goes wrong, so we can fix issues faster. Your data is only visible to the actor developer, not publicly.

Support

Found a bug or have a feature request? Open an issue in the Issues tab on this actor's page. For custom solutions or enterprise integrations, reach out through the Apify platform.