How Healthpeak Transformed Real‑Estate Operations with Palantir AI

Background and Challenges

In traditional commercial real‑estate management, data is siloed in spreadsheets and CRM systems, causing manual data entry, long hand‑off cycles, and poor decision support.

Lease, maintenance, and financial data are scattered.

Cross‑department collaboration relies on manual information transfer.

Historical data is hard to trace, leading to decisions without data backing.

Property managers spend hundreds of hours each month recording meter readings and calculating bills.

Scaling the portfolio linearly increases administrative staff.

Solution Overview

Healthpeak adopted Palantir’s Artificial Intelligence Platform (AIP) and built a four‑layer AI‑driven operating system centered on an ontology.

Physical Layer

Physical assets such as buildings, HVAC equipment, and sub‑meters.

Edge data sources: photos, meter readings, sensor streams.

Data & Ontology Layer

The ontology formally models real‑world entities and relationships.

Property objects – digital twins with location, area, and configuration.

Tenant objects – lease history, space usage, consumption patterns.

Device objects – model, installation date, warranty, maintenance history.

Lease objects – contract terms and billing methods.

Relationship graph linking tenants to spaces, devices to properties, and consumption to tenants.

Intelligence Layer (AIP)

AI agents execute intelligent tasks on top of the ontology.

Computer‑vision OCR extracts meter readings from photos and utility bills.

NLP parses voice recordings from property managers.

Speech‑to‑text converts audio to structured data.

Automated billing engine applies the appropriate billing methodology and generates invoices without human intervention.

Anomaly detection compares current consumption with historical trends and flags outliers for manual review.

Workflow orchestration triggers downstream actions such as notifying leasing teams of expansion opportunities or creating maintenance work orders with diagnostic context.

Interface Layer

Mobile app lets field staff capture data via voice or photos.

Management dashboard provides executives with a global view for capital‑allocation decisions.

Implementation Details – From Data to Insight

Automated Sub‑Meter Billing Workflow

Step 1: Edge data capture – manager photographs sub‑meter; photo uploads to AIP.
Step 2: Computer‑vision – OCR reads the dial, matches device ID in the ontology, calculates incremental consumption.
Step 3: Intelligent billing – query tenant object, retrieve lease billing method, compute charge (e.g., total consumption × tenant area ÷ building area).
Step 4: Anomaly detection – compare month‑over‑month and against forecast; flag >30 % deviation.
Step 5: Invoice generation – system creates and sends invoice, records transaction in financial system.

Voice‑Driven Multi‑Task Workflow

A property manager records a voice note after speaking with a tenant. The system:

Transcribes audio to text.

Extracts entities (building, tenant, expansion request, HVAC issue).

Queries the ontology for current occupancy and growth trends.

Identifies available space and suggests expansion options.

Generates a leasing recommendation and notifies the leasing team.

Looks up the HVAC device, checks maintenance history, and creates a high‑priority work order with diagnostic details.

Technical Benefits

Manual processes reduced from days to hours.

Billing accuracy approaches 100 %.

Property‑manager time spent on tenant interaction rises from ~20 % to 60‑70 %.

Operational cost scales near‑zero as portfolio grows.

AI handles 99 % of routine tasks; humans intervene only on exceptions.

Challenges and Mitigations

OCR errors on diverse utility‑bill formats – addressed with custom vision models, ensemble techniques, confidence scoring, and continuous learning from human corrections.

Legacy CRM/financial systems – integrated via APIs and data import/export; the ontology acts as a semantic middle layer.

Change management – emphasize AI as an augmentation tool, involve staff in design, and demonstrate time‑saving benefits.

IoT device standardisation, real‑time latency, and data‑privacy compliance – identified as ongoing technical hurdles.

Future Vision (2026)

Healthpeak aims for a fully interconnected enterprise operating system where people, buildings, and data are linked through a unified ontology, enabling semantic understanding, intelligent orchestration, and continuous evolution.