In response to the January 2025 wildfires in Los Angeles, AidKit partnered with Los Angeles County and the City of Los Angeles to rapidly launch two emergency cash relief programs.
I led the design and implementation of the platform's new geospatial capabilities for disaster response. This included building a sync pipeline that ingested and normalized data from remote tax parcel datasets and CAL FIRE damage inspection data, storing it in PostGIS for spatial queries. We used this data to automatically triage applicants based on parcel-level impact, reducing manual review and speeding up aid delivery.
I also built real-time, interactive maps to give stakeholders live visibility into program reach and eligibility coverage. These programs ultimately distributed over $25M in direct aid to individuals and small businesses across Los Angeles.
This project inspired me to explore how AI and MCP servers can accelerate program administration. For example, we could connect an LLM to CAL FIRE's damage data to enable a wider range of automated triage and eligibility workflows.
