Fragmented Data
Meter, BEMS, weather, and equipment data remain disconnected.
From Policy to Proof.
HeatAI turns meter data, quality checks, baselines, calculation rules, and change history into reproducible Evidence Bundles for energy and climate-performance verification.
Initial focus: large-scale heat pumps, public buildings, commissioning, and energy-performance M&V.
The Principle
HeatAI does not ask institutions to trust a black-box result. It structures data, rules, calculations, and change history so that third parties can reproduce and review the result.
The Principle
HeatAI does not ask institutions to trust a black-box result. It structures data, rules, calculations, and change history so that third parties can reproduce and review the result.
The Problem
Meter, BEMS, weather, and equipment data remain disconnected.
Baselines and assumptions can change without a traceable record.
Spreadsheet-based calculations are difficult to reproduce and audit.
Final reports show conclusions but rarely preserve the evidence chain.
The missing layer is not another dashboard. It is a reproducible evidence infrastructure.
What HeatAI Does
Operational, meter, BEMS, equipment, and weather data.
Missing values, anomalies, time alignment, and sensor reliability.
Output: QC PASS / WARN / FAIL
Defined reference conditions and calculation assumptions.
Output: Baseline Lock
Equipment, period, variables, and exclusions.
Output: Boundary Seal
SPF, energy savings, emissions reduction, and tCO₂e.
Data, rules, calculations, versions, and change history.
Completeness, quality, consistency, and reproducibility.
Audit-reviewable output for institutions and project stakeholders.
A structured, reviewable record that carries the data, rules, and calculations behind a verification result — not just the conclusion.
Bundle Components
Boundary Seal
Defines exactly what was measured.
QC Certification
Every input validated before it counts.
Versioned Change Log
Nothing changes without a record.
Reproducible Calculation
Rerun the record — same result.
Bundle Properties
Project
Large-Scale Heat Pump Demonstration HP-2026-001
Illustrative sample — not a verified customer result.
Reference Architecture
A layered structure — not a one-off pipeline. Reality enters at the foundation and returns as reproducible evidence that supports the next decision.
Designed to support audit, regulatory review, public investment evaluation, and institutional decision-making. Next: the HeatAI Verification Standard.
Equipment, meters, and field systems captured continuously with their origin intact.
A standardized method resolves operational signals into structured, bounded, quality-controlled records.
The reviewable record — bounded, validated, versioned, and reproducible. The load-bearing center of the architecture.
The point at which structured evidence becomes reviewable by third parties — reproducible and defensible.
Compare verified performance against targets, budgets, and design intent.
A record structured to support audit, regulatory review, and institutional decision-making.
Reviewed evidence returns to the next decision — closing the loop between execution and policy.
Initial Applications
HeatAI starts where operational proof matters most, then extends to broader climate and energy programs.
Verify field SPF, energy performance, operational conditions, and avoided emissions beyond laboratory ratings.
Compare design intent and policy targets with actual operating performance.
Create reproducible evidence for acceptance testing, commissioning, M&V, and performance guarantees.
Expansion Applications
HeatAI Verification Standard
The HeatAI Climate Verification Standard is a proposed specification that defines how operational evidence should be structured, processed, scored, reproduced, and reviewed.
Specification Contents
Standards Alignment
GHG quantification and verification principles
Operational data and calculation evidence structure
Emissions accounting framework
Equipment- and project-level operational evidence
Energy savings measurement and verification
QC, baseline locking, versioning, and reproducibility
Energy management systems
Continuous operational performance evidence
Sustainability disclosure
Underlying operational evidence supporting reported claims
References indicate design alignment only. They do not imply certification, endorsement, or formal conformity.
Verification API
Verification results are structured to flow into institutional systems, disclosures, and audits — not to sit in a static file.
/verificationCreate and run a verification request
/evidence/{id}Retrieve the structured Evidence Bundle
/report/{id}Retrieve the Verification Report
/score/{id}Retrieve the Verification Score and scoring dimensions
Current Status
HeatAI is in active development. The following reflects the honest state of each area — what is built, in progress, under review, or planned.
Core verification workflow and evidence structure developed.
Schema and required evidence components under development.
Six Korean provisional patent applications related to verification and evidence structuring.
Operational data, QC, performance calculation, scoring, and report workflow.
Developing toward institutional review, field validation, and formal standardization.
Pilot Program
Starting price
KRW 8 million
Final scope subject to data and project conditions.
Starting price
Based on project scope
Scope and pricing defined with your team.
Research
HeatAI is developing reproducible methods for operational energy performance, climate-impact calculation, evidence structuring, and machine-readable verification.
Company
Founded in Seoul in 2026, HeatAI develops verification infrastructure that connects real-world energy operations with policy, audit, research, and climate accountability.
Founder & CEO
Jun Yun Lee
20+ years of experience in technology strategy, business development, project management, and digital transformation.
Resources
Documentation and materials are being prepared. Request access and we will share what is available.
Evidence Review
Submit a brief project overview. HeatAI will review the verification objective, available data, boundary, and expected outputs.