Projects | Apex App Solutions

Disclosure Analysis Engine — LLM Parsing for Real Estate Risk

Problem: Real estate disclosure documents often exceed 200 pages, making it difficult for buyers and agents to spot negotiation leverage or structural risks buried in fine print.
Build: I developed an AI tool that parses these massive documents and surfaces critical issues in a clear, organized summary.
Skill demonstrated: Large language model integration for document analysis and information extraction. This tool provides a structured overview to help users navigate complex paperwork and identify key points of concern without the need for manual, line-by-line review.

Problem: Evaluating property or neighborhood quality requires piecing together scattered public records like taxes, crime stats, and school ratings, which is a slow and manual process.
Build: I built a tool that pulls from multiple local data sources into a single snapshot of an area, including core cost and return estimates.
Skill demonstrated: Data aggregation from public sources and automated calculation logic. It simplifies property research by centralizing disparate information into a readable format for quick assessment.

Problem: Traditional radio is fading, and streaming apps lack on-demand voice artist search while driving.
Build: DriveJam (drive-jam.com) is a voice-controlled web app for on-demand playback through car Bluetooth.
Skill demonstrated: Voice interface design, AI-driven search, and real-time audio playback integration.

Problem: Managing specialized diets, food allergies, or small household meal plans is complex, and generic recipe apps rarely account for what's already sitting in someone's kitchen — leading to wasted groceries and daily mealtime stress.
Build: I built Plate IQ (plate-iq.food), an AI-driven meal planning app that generates personalized recipes and grocery lists from a photo of someone's pantry along with their dietary preferences, combining recipe generation with practical shopping list automation.
Skill demonstrated: AI-driven personalization, image-based ingredient recognition, and constraint-based recommendation logic. This tool takes the guesswork out of "what's for dinner" by turning what's already on hand into a usable, personalized plan.