AI-Driven Climate Risk Extraction on Google Cloud
This project implements a serverless, event-driven data pipeline that extracts structured climate adaptation insights for Colombia from IPCC AR6 reports using Google Cloud and Generative AI.
The system automatically transforms unstructured climate science PDFs into structured, analytics-ready datasets.
It demonstrates applied expertise in:
- Cloud-native data engineering
- AI/LLM integration in production pipelines
- Event-driven architectures
- BigQuery data modeling
- Semantic analytics with Looker
Workflow
-
Cloud Storage (IPCC PDFs).
-
Cloud Function Gen2 (Event Trigger).
-
Gemini LLM (Structured Extraction).
-
BigQuery (Curated Dataset).
-
Looker (Semantic Layer).