AVRENIX API

All API requests require a JWT bearer token. Obtain your token from the portal under Settings → API Keys. Tokens are scoped to your plan — Fleet tokens access FoM engines only; Enterprise tokens access all nine engines including PHALANX and TARSIS.

Authorization: Bearer eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9...
Content-Type:   application/json

// Request
{
  "email":    "[email protected]",
  "password": "••••••••••••"
}

// Response
{
  "ok": true,
  "data": {
    "token":      "eyJhbGci...",
    "expires_in": 86400,
    "user": {
      "email": "[email protected]",
      "plan":  "fleet"
    }
  }
}

Runs the complete differentiators cycle: adversarial scenario generation, market microstructure analysis, causal attribution, grid personality update, and autonomous negotiation. Returns a synthesised dispatch recommendation with full engine outputs.

{
  "site_id":              "site-001",
  "battery_power_mw":     100,
  "battery_capacity_mwh": 400,
  "current_soc":          0.75,
  "current_price":         80.00,
  "planned_dispatch_mw":  80,
  "direction":             "discharge",
  "horizon_hours":         4
}

{
  "ok": true,
  "data": {
    "cycle_id":                  "...",
    "recommended_dispatch_mw":   80.0,
    "recommended_direction":     "discharge",
    "confidence":               0.88,
    "estimated_revenue":         /* $ over horizon */,
    "regime":                    "peak",
    "adversarial_report":        { /* AdversarialReport */ },
    "differentiator_insights":  ["Grid regime: PEAK..."]
  }
}

Four-model ensemble (LSTM, XGBoost, Prophet, TFT) with conformal prediction intervals guaranteeing calibrated coverage. Drift detection via ADWIN + PSI alerts when market regime shifts.

{
  "ensemble_forecast": [...],  // forecast horizon values
  "p10":               [...],  // lower prediction interval
  "p90":               [...],  // upper prediction interval
  "mape":              /* error on your data */,
  "coverage":          /* conformal interval coverage */,
  "models": {
    "LSTM":    { "weight": ..., "mape": ... },
    "XGBoost": { "weight": ..., "mape": ... }
  }
}

Bayesian Monte Carlo simulation over stochastic price and load futures. Evaluates dispatch policies (Arbitrage-Opt, Risk-Aware CVaR, RL Aggressive) across thousands of sampled paths and returns Shapley attribution of P10/P50/P90 revenue.

{
  "job_id":              "...",
  "p10_annual_revenue":  /* $ at 10th percentile */,
  "p50_annual_revenue":  /* $ at median */,
  "p90_annual_revenue":  /* $ at 90th percentile */,
  "paths_per_second":    /* observed sim throughput */,
  "shapley": {
    "forecast_accuracy":     ...,
    "regime_classification": ...,
    "soc_management":        ...,
    "ancillary_revenue":     ...
  }
}

Probabilistic peak window forecasting for behind-the-meter facilities. Every candidate dispatch schedule is validated through STRIX before commitment — any schedule with peak miss probability above 5% is rejected. Demand-charge reduction is measured on your own billing data during a pilot.

{
  "peak_risk_probability": /* 0–1 */,
  "window_start":          "14:30",
  "window_end":            "16:00",
  "confidence":            /* 0–1 */,
  "recommended_action":    "discharge_full",
  "demand_charge_liability":/* $ exposure */
}

Joint optimisation across all tariff components — TOU energy charges, demand tiers, NEM export compensation, EV fleet coordination, and 12-month ratchet clauses — as a single objective function. 200+ utility rate structures supported.

{
  "rate_breakdown": {
    "tou_savings":              /* $ */,
    "demand_savings":           /* $ */,
    "export_revenue":           /* $ */,
    "ev_coordination_savings":  /* $ */,
    "total_annual":             /* $ */
  },
  "solve_time_ms":  /* solver wall time */,
  "optimality_gap_pct": /* MILP gap */
}

All error responses follow the same envelope. HTTP status codes are used conventionally. Engine-level errors include the engine name in the error field.

{
  "ok":    false,
  "error": "Invalid SOC value: must be between 0 and 1"
}

// HTTP Status Codes
400  Bad Request — validation error in request body
401  Unauthorized — missing or invalid token
403  Forbidden — plan does not include this engine
422  Unprocessable — parameter range violation
429  Rate limited — see X-RateLimit-Reset header
500  Engine error — see error field for detail

Rate limit headers are included on every response: X-RateLimit-Remaining and X-RateLimit-Reset. The Operator and Fleet figures are the target per-plan limits; Enterprise limits are set in the partner agreement rather than fixed here. These are the intended allocations, scoped per partner — not a validated throughput guarantee at scale.