Platform

Predict. Detect. Prove.

The energy intelligence platform that turns building data into forward-looking insight, validated anomalies, and audit-ready evidence — across every site in your portfolio.

Built on Temporal Fusion Transformer
Five peer-reviewed papers
REST API + Python & TypeScript SDKs
app.sensecoreai.com/dashboard/forecasting
Forecasting · Civic Centre HQ
Next 48 hours · Half-hourly resolution
LIVE
Actual
Forecast
95% confidence
MODEL
TFT v1.2
MAPE
10.31%
WEATHER
INTEGRATED
EXPLAIN
SHAP READY
What's inside

Eight capabilities. One platform. One API.

Forecasting
TFT model with weather covariates and 10.31% MAPE benchmark.
Anomaly detection
Three-method ensemble: Z-Score, IQR and Isolation Forest.
Carbon reporting
Automated Scope 1 + 2 using DEFRA conversion factors.
Compliance
SECR, ESOS, ERIC, SHDF and EPC band tracking in one place.
Fuel Poverty Score
Proprietary tenant-level FPRS for regulated social housing.
Cost optimisation
Tariff comparison and savings engine surface lost spend.
Explainability
SHAP attribution for every prediction the platform makes.
Integrations
CSV upload, smart meter feeds, REST API, Python and TS SDKs.
Forecasting

Forecast demand with research-grade accuracy.

SenseCore's Temporal Fusion Transformer model forecasts electricity, gas, solar and water consumption with industry-leading precision. Trained on the BDG2 benchmark — the same open dataset used by leading energy research — and validated through five peer-reviewed papers published on Zenodo.

  • 48-hour, 7-day, and 30-day forecast horizons
  • Weather covariates from Open-Meteo, refreshed automatically
  • Occupancy inference via Gaussian Mixture Models
  • SHAP feature attribution for every prediction
  • Plug-in architecture: gas, solar, water deploy as toggles
Explainability · SHAP attribution
Why this prediction
Civic Centre HQ · Next 24h forecast · +44.7 kWh
Feature importance
Increases
Decreases
Hour of day
+18.2 kWh
Day of week
+12.4 kWh
Outdoor temperature
+9.1 kWh
Recent consumption
+6.8 kWh
Occupancy estimate
+6.2 kWh
Heating degree-days
−4.5 kWh
Holiday flag
−2.1 kWh
Solar irradiance
−1.4 kWh
Model TFT v1.2
Method SHAP
Retrained 14 days ago
Anomalies · Last 24 hours
Detected anomalies
3 OPEN
SEVERE
HVAC scheduling drift
£425/yr
Council Office HQ
Z-Score
IQR
MODERATE
Peak demand exceeded
£180/yr
East Wing School
Isolation Forest
MINOR
Off-hours consumption spike
£60/yr
Library HQ
Z-Score
Anomaly Detection

Catch waste before it scales.

Three independent detection methods running in parallel surface anomalies that single-model systems miss. Each alert is severity-scored, cost-impacted, and SHAP-explained — so your team knows exactly which signals deserve action and which can be safely ignored.

  • Z-Score baseline for statistical outliers
  • IQR thresholds robust to non-normal distributions
  • Isolation Forest for complex multivariate patterns
  • Estimated cost impact attached to every anomaly
  • Per-building threshold configuration for tuning
Compliance & Carbon

Reporting that's already audit-ready.

Six UK regulatory frameworks automated from one API. Generate SECR Scope 1 + 2 evidence, ESOS Phase 4 audits, NHS ERIC returns, SHDF retrofit scoring, and EPC band tracking — versioned, signed, and exportable as branded PDF reports.

  • SECR — Scope 1 & 2 with current DEFRA factors
  • ESOS — Phase 4 evidence pack generation
  • NHS ERIC — energy return automation
  • SHDF — retrofit readiness scoring
  • EPC — band tracking and projection
  • Branded PDF exports, audit-versioned
Compliance status · 2026
Reporting workflows
Auto-updated
SECR Scope 1 + 2
ON TRACK
94%
Due Mar 2027
ESOS Phase 4
ON TRACK
78%
Due Dec 2027
NHS ERIC return
COMPLETE
100%
Submitted
SHDF readiness
IN PROGRESS
65%
Due Jun 2026
EPC band tracking
ACTIVE
88%
Continuous
Fuel Poverty Risk Score
Portfolio risk distribution
12
HIGH
34
MEDIUM
178
LOW
412
CLEARED
Risk factor signals
EPC band
78
Consumption
62
Household
45
Intervention
30
Fuel Poverty Risk Score

The only platform with FPRS built in.

A proprietary Fuel Poverty Risk Score gives social housing teams tenant-level visibility into properties at risk — combining EPC band, consumption patterns, household indicators and intervention history into one actionable score that scales across thousands of properties.

  • Tenant-level risk scoring across whole portfolios
  • EPC band signals integrated automatically
  • Consumption-derived risk indicators
  • Retrofit prioritisation for SHDF wave readiness
  • Designed for regulated social housing providers
Integrations & API

Connect data once. Use it everywhere.

SenseCore ingests data from smart meters, CSV uploads and direct API push — and outputs forecasts, alerts and reports through your dashboard, REST API, Python SDK or TypeScript SDK. Hybrid authentication supports both interactive sessions and machine-to-machine integration.

Data in
  • Smart meter feeds (via N3rgy)
  • CSV and Excel upload
  • REST API push
  • Half-hourly resolution
Data out
  • REST API (v1, X-API-Key auth)
  • Python SDK (PyPI-ready)
  • TypeScript SDK (npm-ready)
  • Webhooks with retry backoff
~/projects/energy — zsh
# Forecast next 48 hours for a building
$ curl -X POST https://api.sensecoreai.com/api/v1/forecast \
-H "X-API-Key: sk_live_abc..." \
-d '{"building_id":"abc-123","horizon_hours":48}'
# Response
{
"building_id": "abc-123",
"horizon": "48h",
"forecast": [...96 values],
"mape_estimate": 10.31,
"model_version": "tft_v1.2.1",
"explainability": "shap_available"
}
Built for trust

Architecture and security designed for procurement.

AWS UK hosting
eu-west-2 deployment with UK data residency. Daily encrypted S3 backups.
ISO-aligned controls
Encryption at rest and in transit. Access scoped per organisation.
Role-based access
Per-org permission model. Viewer, analyst, admin and super admin roles.
Audit logging
Database-backed request log. Every API call traced and stored.
What you can build

Three things teams ship in the first month.

Energy intelligence isn't abstract — here's what real workflows look like once SenseCore is connected to your meter data.

01
A live forecast dashboard
Pull tomorrow's consumption per building, with confidence bands and weather covariates baked in. Replaces a Tuesday-morning spreadsheet that took hours.
For: facilities managers, energy buyers
02
An auto-generated SECR report
Connect data once, get versioned SECR Scope 1+2 reports with DEFRA factors applied. SHA-256 hash on every output for audit traceability.
For: compliance leads, finance teams
03
Fuel-poverty triage list
FPRS ranks every tenant property by risk and shows the four signals driving each score. Triage support outreach to the households that need it most.
For: housing officers, sustainability leads
See it on your own building data.
Book a 30-minute demo and we'll show you a forecast, an anomaly scan, and a compliance run using a sample of your portfolio.
Book a demo →