Real Estate AI Data Agent
Modern Data Ecosystem for National Newspaper
Overview
Led distributed team to transform a national newspaper's underutilized 210GB+ real estate data into a highly efficient AI Data Agent, delivering real-time business signals and actionable insights through custom APIs and modern data infrastructure.
Challenge
A newspaper organization was sitting on a wealth of data but lacked technical expertise and infrastructure to organize, manipulate, and extract meaningful insights. They had no strategy for managing data, no API structure for current or future projects, and needed to transform raw data into usable business signals accessible to development teams.
My Role
Senior Technical Product Manager - Led distributed team, developed comprehensive data strategy, facilitated discovery workshops with 23 stakeholders across 7 departments, managed API-first architecture, data governance, CI/CD, and defined roadmaps for multiple teams.
Approach
Leveraged Snowflake and dbt to clean and structure existing data with continuous updates. Built two key APIs: (1) Real-time Kafka event-streaming API for business transactions and property records tied to individuals/businesses, (2) Poll API for detailed query-based access to specific records. Organized data into categories connecting people, businesses, and properties to generate actionable business signals.