What was the impact of the AI solution?

Output accuracy increased from under 95% to 99.9%. Manual processing dropped from roughly four hours to around a minute and in some cases to 10 seconds. Brokers save 4 to 6 hours of manual work per object. Gazelle launched four new products in less than a year.

Source

GCPEvidence: Medium50/100

Gazelle automates real-estate document extraction and content generation with Gemini 1.5 Pro

Use case typeDocument processingUpdated Jun 13, 2026

Gazelle, a Sweden-based real estate brokerage automation company, integrated Gemini 1.5 Pro into broker workflows to extract key information from long property documents and generate property descriptions and summaries in the desired format. The solution uses APIs, Gemini multimodal capabilities, BigQuery for performance analytics, and Google Maps Platform plus Google Search context for grounded area descriptions.

Organization: Gazelle
Industry: Real Estate
Location: Sweden
Published: January 2024

Reported outcomes

95-99.9%

accuracyQuality & accuracy

10 secondstime4-6 hourstime

Strategic outcomes

New product / capabilityAutomated property description generationNew product / capabilityExtracted key financial document dataSpeed & agilityReduced manual document processingNew product / capabilityLaunched four new products

Catalog median for quality & accuracy deployments: +41% across 63 reported metrics. Compare benchmarks →

Why do we believe this?Outcome claims, sources, and evidence checks

Normalized claim

Accuracy: 95-99.9% increase

Google Cloud Customer StoryJan 1, 2024Customer storyInferred claimMedium evidence strength

Output accuracy increased from under 95% to 99.9%.

Normalized claim

Time: 10 seconds decrease

Google Cloud Customer StoryJan 1, 2024Customer storyInferred claimMedium evidence strength

Manual processing dropped from roughly four hours to around a minute and in some cases to 10 seconds.

Normalized claim

Time: 4-6 hours decrease

Google Cloud Customer StoryJan 1, 2024Customer storyInferred claimMedium evidence strength

Brokers save 4 to 6 hours of manual work per object.

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Gazelle, EiendomMegler 1 Oslo Akershus AS
Provider: GCP
Maturity: Unknown
Linked source: Google Cloud Customer Story

No explicit deployment-stage evidence found.

Customer identity supportedSource describes one deploymentMaturity evidence evaluated

Primary read

Use case focus

Showing 3 of 3

1Document Processing
2Content Generation
3Workflow Automation

Automate property description generation and extraction of key financial data from complex building-inspection and electricity documents without hallucinated or inaccurate outputs.

Gazelle integrated Gemini into brokers’ workflows via APIs to process lengthy documents, extract information, and generate summaries and descriptions.
The company uses BigQuery to analyze performance and Google Maps Platform plus Google Search grounding for area-description generation.

Technologies

Gemini 1.5 Pro BigQuery Google Maps Platform Google Search

Output accuracy increased from under 95% to 99.9%.
Manual processing dropped from roughly four hours to around a minute and in some cases to 10 seconds.
Brokers save 4 to 6 hours of manual work per object.
Gazelle launched four new products in less than a year.

Architecture

Gazelle integrated Gemini 1.5 Pro into broker-facing workflows via APIs to extract data from complex property documents and generate text outputs. BigQuery is used to analyze tool performance, and Google Maps Platform plus Google Search are used to ground future area-description generation.

Sources & evidence1

Evidence: Medium50/100Evidence strength