ROI = Real Objects Intelligence: actionable, timestamped intelligence objects — not stale datasets.
One query. Six subsystems. Full intelligence package.
19 AI skills, semantic memory, 50 state experts, event machine.
Traditional CRE: 200 cold calls/day, manual comp pulls, hand-written IC memos.
SwarmCRE: one query → full intelligence package → 2M dials/day at edge speed.
Natural language or structured. Heuristic parser extracts asset type, state, city, SF, price, cap rate, intent in <5ms.
6 subsystems run in parallel: semantic memory, event database, entity resolution, market context, infrastructure map, state expert.
Full package in <2s. Market heat, comps, events, entities, skill recommendations. Deep mode executes skills for you.
Each skill: system prompt → Qwen3-30B edge AI → JSON validation → R2 finality.
Execute individually or let the search orchestrator recommend the right ones.
Chain API calls into real CRE workflows. Each step feeds the next. All endpoints work on the free router domain.
Full deal analysis from a single property description.
POST /search
Get parsed query + market context + skill recommendations
POST /skill/broker_senior
Deal verdict, pricing, risk assessment
POST /skill/debt_analyzer
DSCR, LTV, debt terms from step 2 pricing
POST /skill/comp_analyzer
Comparable sales and market positioning
Deep market analysis with events, memory, and trends.
POST /search {"deep": true}
Full intelligence package + auto skill execution
GET /events/query?type=just_listed&state=TX
Recent deal activity in the target market
POST /memory/search
Historical intelligence objects and trends
Exchange analysis, site selection, and tax impact.
POST /skill/exchange_1031
Timeline, boot calc, replacement requirements
POST /skill/site_selector
Target markets matching exchange criteria
POST /skill/tax_assessor
Tax implications per target state
One endpoint. Full intelligence package. No SDK required.
router.swarmandbee.com = public/basic. api.router.swarmandbee.com = authenticated + metered.
curl -X POST https://router.swarmandbee.com/search \
-H "Content-Type: application/json" \
-d '{"query": "85K SF warehouse Alliance TX 6% cap"}'
# Deep mode (executes recommended skills)
curl -X POST https://api.router.swarmandbee.com/search \
-H "Content-Type: application/json" \
-H "X-API-Key: sb_live_your_key" \
-d '{"query": "cold storage Miami FL", "deep": true}'import requests
# Basic search (free on router domain)
r = requests.post("https://router.swarmandbee.com/search",
json={"query": "85K SF warehouse Alliance TX 6% cap"})
data = r.json()
print(f"Parsed: {data['parsed']}")
print(f"Market: {data['market_context']['label']}")
print(f"Skills: {data['skills']['recommended']}")
print(f"Latency: {data['latency_ms']}ms")const res = await fetch("https://router.swarmandbee.com/search", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({
query: "85K SF warehouse Alliance TX 6% cap"
})
});
const data = await res.json();
console.log(`Parsed: ${JSON.stringify(data.parsed)}`);
console.log(`Market: ${data.market_context?.label}`);
console.log(`Skills: ${data.skills.recommended}`);
console.log(`Latency: ${data.latency_ms}ms`);From zero to your first Intelligence Package in 3 steps.
No API key needed — the public router is free.
One cURL command. Hits the search orchestrator. Returns a full intelligence package.
curl -X POST https://router.swarmandbee.com/search \
-H "Content-Type: application/json" \
-d '{"query": "85K SF warehouse Alliance TX 6% cap"}'Paste in your terminal. Or click below to run it right here.
Parsed query, market context, infrastructure, memory matches, recent events, entity graph, skill recommendations. All in <2 seconds.
Not text. Not summaries. Not chat. A bankable intelligence package.
Every object: schema-validated, machine-usable, composable, auditable, Hedera-timestamped.
Scenario: "950K SF single-tenant industrial Florida" → full package in <2s
{
"object_type": "AssetIntelligenceObject",
"asset_type": "industrial_single_tenant",
"state": "FL",
"sf": 950000,
"clear_height_ft": 36,
"dock_doors": 80,
"trailer_parking": 200,
"rail_served": false,
"sprinkler": "ESFR",
"zoning": "industrial",
"occupancy": 1.0,
"lease_type": "NNN",
"market_rent_psf": 8.25,
"vacancy_market": 0.042,
"hedera_timestamp": "2025-02-24T18:32:01.443Z",
"confidence": 0.94,
"source_skill": "broker_senior"
}{
"object_type": "MarketSnapshotObject",
"market": "Central Florida",
"asset_class": "industrial",
"rent_range_nnn_psf": [6.50, 9.75],
"vacancy_pct": 4.2,
"cap_rate_range": [5.25, 6.25],
"market_heat": 82,
"tier": 1,
"population_growth_yoy": 0.019,
"job_growth_yoy": 0.027,
"major_employers": [
"Amazon", "FedEx", "Publix", "Chewy"
],
"infrastructure": {
"interstate": ["I-4", "I-75", "I-95"],
"port": "Port Tampa Bay / Port Everglades",
"airport": "MCO / TPA"
},
"tax_environment": {
"property_tax_rate": 0.018,
"income_tax": false,
"incentive_zones": ["Enterprise Zone", "QOZ"]
},
"hedera_timestamp": "2025-02-24T18:32:01.443Z",
"confidence": 0.91,
"source_skill": "market_report"
}{
"object_type": "FinancialObject",
"base_rent_psf": 8.25,
"gross_income": 7837500,
"stabilized_noi_estimate": 7700000,
"lease_structure": "NNN",
"tenant_pays": ["taxes", "insurance", "CAM"],
"landlord_expense": "asset_mgmt + reserves",
"confidence_score": 0.74
}950K SF × $8.25 NNN = $7.84M gross. Single-tenant NNN: tenant pays taxes, insurance, CAM. NOI ≈ $7.7M after minor reserves.
{
"object_type": "ValuationObject",
"low_estimate": 123200000,
"high_estimate": 146600000,
"midpoint": 135000000,
"value_per_sf_mid": 142,
"cap_rate_low": 0.0525,
"cap_rate_high": 0.0625,
"noi_used": 7700000,
"confidence_score": 0.72
}At 5.25% cap: $146.6M. At 6.25% cap: $123.2M. Midpoint ≈ $135M ($142/SF).
{
"object_type": "RiskFlagObject",
"tenant_concentration": "100%",
"lease_term_dependency": "Critical",
"re_leasing_risk": "High if specialized build",
"insurance_exposure": "Elevated in FL",
"storm_exposure": "Location-specific",
"market_oversupply_risk": "Monitor deliveries pipeline",
"confidence_score": 0.77
}{
"object_type": "AdvisoryObject",
"investment_profile": "Core / Core+",
"strength_score": 0.82,
"risk_score": 0.61,
"overall_confidence": 0.75,
"strengths": [
"Long-term NNN lease",
"Strong FL logistics tailwinds",
"Institutional-grade bulk size"
],
"watch_items": [
"Lease term remaining (years left?)",
"Tenant credit rating",
"Re-leasing feasibility if vacated"
],
"executive_summary": "A 950K SF stabilized single-tenant industrial asset in Florida represents an institutional-grade logistics investment with estimated valuation between $123M and $147M depending on tenant credit, lease term remaining, and submarket positioning."
}{
"object_type": "ActionPlanObject",
"next_steps": [
"verify lease term remaining",
"verify tenant credit rating",
"pull 900K+ SF comps (FL industrial)",
"stress test cap rate expansion +100-150bps",
"analyze insurance trend year-over-year"
],
"skills_to_execute": [
"comp_analyzer",
"debt_analyzer",
"tax_assessor",
"investor"
]
}7 objects. Machine-usable. Auditable. Composable. Final.
This is not a dataset row. This is a bankable intelligence package.
Pay for what you use. Basic search is free on the router domain.
Every Intelligence Object timestamped on Hedera Consensus Service.
HCS-10 OpenConvAI compliant. Agent-to-agent communication on public ledger.
Every ROI object is auditable and replayable — receipts for agents.
Peer-to-peer agent messaging via Hedera Consensus Service. Connection management, registry discovery, signed transactions.
Agent identity on-chain. 19 skills, capabilities, endpoints — discoverable by any agent in the OpenConvAI ecosystem.
Every cooked object stored in R2 with Hedera consensus timestamp. Immutable provenance for CRE intelligence.
0.0.102988340.0.102988370.0.10298836HCS-10 OpenConvAI