Off-Market Property Intelligence Platform for the Canadian Market

FirstList (formerly Knok'd) is a real estate technology platform that helps buyers, agents, and investors discover off-market property opportunities before they appear on traditional listing channels. By expanding access to exclusive inventory, the platform enables users to identify opportunities that are often missed through conventional property search methods. 

As demand for timely property intelligence increased, expanding beyond traditional listing sources became an important part of the platform's growth strategy. Reliable access to alternative property data helped strengthen inventory quality while improving the speed at which new opportunities became available to users.

Many off-market properties are shared within Facebook communities using free-form descriptions, images, and inconsistent posting formats. Although these communities contain valuable opportunities, the information is difficult to organize into a consistent and searchable dataset that can support real estate workflows. 

Key Issues 

• Monitoring multiple Facebook communities for newly published property opportunities  
• Interpreting inconsistent listing formats created by different users  
• Converting unstructured content into standardized property records  
• Managing listing images with extracted property information  
• Delivering validated listing data into downstream platform workflows  
• Scaling property discovery without increasing manual operational effort  

Manual collection required continuous review of Facebook communities, delaying listing availability, and making it increasingly difficult to expand coverage as more property sources were added.

An AI-powered property intelligence pipeline was developed to automate the discovery, processing, and delivery of off-market real estate listings. The solution transformed unstructured Facebook content into reliable property intelligence that could be consumed directly by the FirstList platform. 

Automated Property Discovery 

The platform continuously monitored targeted Facebook communities using Apify to identify newly published property opportunities. Posts and associated images were collected automatically, eliminating the need for manual monitoring of multiple groups.  

AI-Powered Listing Intelligence 

OpenAI analyzed property posts to identify and organize important listing information into a standardized structure. This transformed inconsistent social media content into reliable property records suitable for downstream business workflows.  

Centralized Listing Management 

Structured property records, listing images, and extracted metadata were stored using PostgreSQL, Prisma, and Supabase Storage. The pipeline generated normalized datasets that could be consumed by application services. This created a consistent foundation for search, listing management, and future platform capabilities.  

Product Integration Framework 

Validated listing records were transformed into API-ready payloads and delivered to downstream services through an automated ingestion workflow. The modular architecture simplified future enhancements while supporting continuous property discovery as the platform expanded.  

Key Deliverables 

• Built an AI-powered property listing intelligence pipeline  
• Automated Facebook property discovery using Apify  
• Converted unstructured property posts into standardized listing records with OpenAI  
• Centralized property images and metadata using PostgreSQL and Supabase Storage  
• Generated normalized payloads for downstream platform integration  
• Developed an API-driven ingestion workflow for continuous listing updates  
• Built a scalable architecture for expanding property intelligence across additional data sources  

Faster Property Discovery 

Automated monitoring of Facebook communities reduced the time required to identify off-market property opportunities, allowing new listings to reach the platform more quickly.  

Improved Listing Quality 

Standardized property information into a consistent structure, making listings easier to search, compare, and use across product workflows.  

Reliable Property Intelligence 

Centralized listing records, images, and metadata into a unified data pipeline, providing dependable property information for downstream application features.  

Scalable Data Collection 

Established an automated acquisition framework capable of supporting continuous property discovery while simplifying future expansion across additional real estate data sources.