DEVELOPERS

Water chemistry as an API.

If you are a water treatment service company white-labeling a customer portal — or an OEM or software team building a product that needs real water chemistry — you can license ours instead of spending years building a solver.

Request an API key Talk about embedding

DEPLOYMENT FOR INTEGRATORS

Our cloud, your cloud, or inside your product footprint.

If you white-label a customer portal or ship software that runs behind a firewall, the engine ships with it. On-premise and air-gapped deployments use the same chemistry engine and identical results against a pinned database version. No feature tier is reserved for cloud-only integrators.

Our cloud (SaaS API)Your cloud / VPCOn-premise in your productAir-gapped capable

Pinned chemistry database versions keep your product’s outputs reproducible across releases
Versioned database bundles support long-term auditability for your customers
Identical results across hosted and embedded footprints against the same database version

Deployment and security details →

INTEGRATION MODELS

Three ways to embed water chemistry

White-label your customer portal, call the API, or embed the engine. Availability is stated honestly — contact us for the current contract path for each model.

REST API

Per-call metered access to speciation, saturation, kinetics, corrosion, and water-report endpoints. Suitable for service-company portals, SaaS products, and multi-tenant backends.

Available — contact us

Embedded engine license

In-process, nonexclusive, metered license for products that cannot pay round-trip latency on every solve. Same chemistry database versioning and provenance shape as the API.

Full platform white-label

White-label your customer portal: run AquaAdvisor’s modeling, twin, and program surfaces under your brand. Also used for OEM suites.

ENDPOINT REFERENCE

API teasers by domain

Paths shown are API teasers for integrator discussions. Exact public paths, versioning, and GA status are confirmed per engagement — ask when you request a key. AI insight and report endpoints are listed for discussion; availability is not claimed as generally available until confirmed.

Speciation & saturation

POST

/api/v1/measured/speciation/point

Single-point aqueous speciation and activity

POST

/api/v1/measured/speciation/batch

Vectorized batch speciation for scenario sweeps

GET

/api/v1/measured/saturation/salt

Saturation index for a named salt / mineral

Kinetics

POST

/api/v1/kinetics/precip

Precipitation / deposition kinetics solve

GET

/api/v1/kinetics/params

Kinetic parameter lookup for a mineral set

Corrosion

POST

/api/v1/measured/corrosion

Mixed-potential corrosion rate (Faraday / mpy)

POST

/api/v1/measured/corrosion/pitting

Pitting electrochemistry screen

Water reports

POST

/api/v1/water/extract

Extract structured chemistry from a lab report

POST

/api/v1/water/analyze-data

Analyze structured water chemistry payload

Compliance

POST

/api/v1/compliance/profile

Regulatory profile for a site or jurisdiction pack

POST

/api/v1/compliance/monitoring-spec

Monitoring specification generation

GET

/api/v1/compliance/status

Compliance status against active monitoring plan

Solver

POST

/solver/solve

Direct solver entry — tier, residual, and provenance on result

AI insights & reports

POST

/api/v1/ai/insights

Generate physics-grounded insights from solved chemistry (path and GA status — contact us to confirm)

POST

/api/v1/ai/reports

Program / exception report generation from model outputs (path and GA status — contact us to confirm)

AI VIA API

Insight and report generation over the API

The AI layer consumes solver and twin output — the same physics floor as the product UI. Whether insight and report generation are exposed as public API endpoints, and under which paths, is confirmed per engagement. Do not treat the teaser paths above as a published contract.

How the AI layer works →

REQUEST / RESPONSE

Every result carries its proof.

The response shape is a selling point: value, thermodynamic database id, activity model, residual, solver tier, and provenance. An accuracy claim without those fields is marketing, not engineering.

curl -s -X POST https://api.aquaadvisor.ai/api/v1/measured/speciation/point \
  -H "Authorization: Bearer $AQUAADVISOR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "temperature_C": 40,
    "composition": {
      "Ca": 120,
      "Mg": 35,
      "HCO3": 180,
      "SO4": 90,
      "Cl": 60,
      "SiO2": 25,
      "pH": 8.2
    },
    "activity_model": "b-dot",
    "database": "aa-thermo-2026.1"
  }'

Response

{
  "ok": true,
  "result": {
    "values": {
      "SI_calcite": 1.42,
      "CCPP_mg_L": 48.6,
      "I_molal": 0.018
    },
    "database": "aa-thermo-2026.1",
    "activity_model": "b-dot",
    "residual": 2.1e-12,
    "tier": "full-registry",
    "provenance": {
      "engine": "aquaadvisor-chem",
      "engine_version": "2026.1.0",
      "solved_at": "2026-08-21T14:02:11Z",
      "temperature_C": 40,
      "composition_hash": "sha256:…"
    }
  }
}

Illustrative model output. Not customer data.

ENGINEERING CHARACTERISTICS

Built for modeling workloads — not a one-shot desktop calc

Vectorized batch endpoints for scenario sweeps and twin ensembles
Differentiable outputs for sensitivity analysis and optimization
Deterministic and reproducible given a pinned chemistry database version
Explicit versioning of chemistry databases so results stay auditable years later
Queued workers for heavy jobs — describe architecture, not unpublished latency targets

Latency and throughput figures are published only as measured, versioned benchmarks on the chemistry engine page — never as marketing estimates.

Request an API key or talk embedding.

Tell us whether you need a white-label customer portal, metered REST, an in-process license, or on-premise inside your product. An engineer will respond with the honest availability picture.

Request an API key Talk about embedding