Revealio: AI-Powered Property Condition Scoring
Created by: Jena Peoples & Bryce Masterson
Real-estate investors traditionally spend hours “driving for dollars” — searching for run-down properties that signal potential opportunities. Revealio replaces that entire process with an automated, AI-first pipeline that can evaluate entire cities in minutes.
What It Does
Revealio collects property-specific imagery from Google Street View, analyzes every exterior with computer vision and dual LLM grading systems, and assigns each home a “distress score.” It then merges that score with ownership data, public records, and geospatial information to produce a refined, high-value lead list for investors.
How It Works
Akka, Redpanda, Neo4j, OpenAI, Anthropic, Google Maps API, CoreLogic, Python, Docker, Loveable.
Revealio impressed the judges with its clean architecture, strong real-estate insight, and near-production-ready pipeline.
Integrates with Google Maps / Street View to capture images at scale
Uses a multi-agent LLM system (OpenAI + Anthropic) where one model detects distress indicators, and another evaluates the opposite, reducing bias
Aggregates data from CoreLogic and other sources
Uses Akka to orchestrate distributed, fault-tolerant workflows across all agents and data-processing components
Processes entire markets in minutes thanks to a distributed architecture