ZoningPal
AI powered zoning intelligence for Toronto providing instant, detailed zoning briefs
Overview
Zoning research takes days of manual work. Municipal bylaws are inconsistent, buried in PDFs, and full of legal language that references other bylaws. Architects spend hours trying to figure out setbacks, height limits, and parking requirements for each project. ZoningPal automates this. It parses Toronto's zoning bylaws, extracts the relevant regulations for any property, and generates clause cited briefs in minutes. We're not replacing expertise. We're giving professionals their time back so they can focus on design instead of document hunting.
The Challenge
The technical challenge is that zoning regulations weren't designed to be machine readable. Each municipality structures their bylaws differently. Toronto's zoning data exists across PDFs, HTML tables, and nested conditional logic. Overlays can modify base zoning rules. Some regulations reference other regulations. Building a system that accurately extracts and interprets this requires both NLP capabilities and deep domain knowledge of how zoning actually works in practice.
The Solution
We built this using Node.js and Claude API to parse Toronto's zoning bylaws, extracting setbacks, density limits, use permissions, and all the relevant clauses. The system handles inconsistent document formats (PDFs, HTML, nested tables) and outputs structured, citation-backed briefs that architects can hand directly to clients. Despite the technical complexity on the backend, the user experience is dead simple: enter an address, get a comprehensive zoning brief in minutes. I co-led the technical development, building the full stack application with React/TypeScript frontend, Node.js/Express backend, and PostgreSQL with PostGIS for spatial queries. Integrated Claude API for regulatory text analysis and Puppeteer for PDF generation. The hardest part wasn't the AI. It was handling edge cases like overlays, conditional clauses, and zone specific exemptions while keeping the interface intuitive. Most competitors scrape planning data or offer manual research services. ZoningPal actually reads the regulations and tells you what applies to your property.
Impact
Co-founded with Alireza and launched to paying customers. Presented technical demos at AI Tinkerers Toronto and Innovate Toronto in 2025. Showcased at Collision 2024 as part of the ALPHA startup exhibition. The platform is live and processing payments, proving there's real demand for tools that automate regulatory research in real estate and architecture.