350,000+
Peer-reviewed articles in the evidence corpus
22
Cities monitored using the EHSPI composite environmental health index across 5 dimensions
7
Google Environmental APIs ingested per tile per refresh
5
Chronic condition sensitivities with evidence-based scoring
49
Raw and computed global environmental data points

70–80% of chronic disease burden is attributable to environmental exposures. No platform connects this to clinical care.

The exposome — the totality of environmental exposures across a lifetime — is now a recognised field of study, pioneered by Christopher Wild and pursued by institutions including Exposome-NL at Utrecht University and ISGlobal in Barcelona. The burden is not disputed science.

Despite this, no platform exists that connects real-time environmental conditions to individual condition management with peer-reviewed scientific evidence. Existing tools provide air quality numbers. They do not connect those numbers to specific conditions, specific populations, or specific evidence.

vasus.ai closes this gap. For each query, the platform ingests real-time environmental data across seven signal types, retrieves evidence from 350,000+ papers, applies condition-specific weighting, and synthesises a structured output with inline citations — delivered via API.

Four steps from raw environment
to cited intelligence

01
🌍
Environmental ingestion

AQ, pollen, barometric pressure, temperature, humidity, UV, and heat index — seven signal types from Google Environmental APIs. Updated hourly across 22 global cities.

02
⚖️
Condition-specific signals

Seven environmental signals weighted by clinical evidence for each of five sensitivities. Barometric pressure delta: 38% weight for migraine. Pollen exposure load: 40% weight for allergies.

03
🔬
Evidence retrieval

Semantic search across 350,000+ peer-reviewed papers via pgvector HNSW index. LLM gate filters for relevance. Top papers ranked and selected. BAAI/bge-base-en-v1.5 embeddings.

04
📊
Intelligence output

Risk level, environmental context, cited recommendations, and uncertainty notes — structured JSON delivered via API. Every response is traceable to peer-reviewed evidence.

This is what vasus.ai returns.

Every response includes the environmental reading, the clinical mechanism, and the peer-reviewed evidence. Not a prediction — an evidence-based intelligence output.

“Barometric pressure dropped 6.2 mb in the last 24 hours — consistent with trigger conditions in barometric migraine literature. Pressure volatility elevated across the 72h window. Recommend preventative review if medication is available.”
Environmental breakdown — London · Migraine · 24h forecast
Barometric pressure delta (24h)
↓ 6.2 mb HIGH
Pressure volatility (72h std dev)
↑ 8.4 mb HIGH
Peak temperature (6h, 90th pct)
17.2°C
Heat index mean (72h)
16.3°C
PM2.5 mean (72h)
✓ 8.1 µg/m³ LOW
"risk_level": "Moderate",
"ehspi_composite": 62,
"top_signals": ["pressure_delta", "volatility"],
"citations": [2 papers],
"condition": "migraine"
EHSPI v1.6 · Updated nightly

A composite environmental health index. Condition-specific. Transparent. Live.

The Environmental Health Sensitivity Performance Index scores locations 0–100 across five chronic condition dimensions. Unlike generic AQI, each sensitivity uses clinically-grounded signal weights derived from peer-reviewed literature. All five weight vectors are published transparently — no other commercial environmental health platform does this.

Migraine sensitivity · Weight vectors Σ = 1.00
Barometric pressure delta (24h abs mean)
0.38
Pressure volatility (std dev, 72h)
0.28
Peak temperature (90th pct, 6h)
0.12
Heat index mean (72h)
0.10
Dew point mean
0.06
AQI peak & PM2.5 mean
0.06